Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.
InPhase Senior Member Joined: 15/12/2020 Location: United StatesPosts: 178
Posted: 05:57am 15 Dec 2020
Copy link to clipboard
Print this post
I'm pretty much a noob when it comes to practical inverter knowledge. I get the theory and am familiar with many inverters from an installation and use view point, but the specific engineering challenges of inverter construction is a field I am only just now wading into.
I have a 12 V to 170 V boost converter I made from the high voltage DC push-pull section of a MSW inverter. The inverter it came from was rated at 3000 watts continuous, which is pretty insane for a 12 volt system, but nonetheless it will do every bit of it according to my tests. The limiting factor there being the battery bank size.
It occured to me that perhaps I could use a sine wave inverter driver (e.g. egs002) to make AC directly at that level, leaving the step up work to the DC converter. But wait! Once that is done, could the output then be fed into a 1:1 transformer to add some "magnetic inertia" for handling inductive loads, etc.?
In other words, would a high frequency switching DC boost supply feeding a transformer at low frequency behave like a LF inverter fed directly from a battery?
Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406
Posted: 07:16am 15 Dec 2020
Copy link to clipboard
Print this post
Three kilowatts sounds like a lot but it really isn't if you are running things that have a very high starting inrush current.
That is the problem with things such as high frequency boost converters, they work fine with constant loads, but they have zero extra surge capacity. They tend to go *BANG* and let the smoke out fairly easily.
You would be much better with a low frequency transformer inverter that runs straight off 12v and drives a 60Hz step up transformer to 110 volts. A big steel transformer can easily handle a x10 overload for several seconds and not raise a sweat.
As long as you have enough mosfets in your inverter with a big enough heatsinks, that too can handle a massive short term burst of heat. It will be big, heavy, and clunky, but it will also be bullet proof.
That is what we mostly do here, multi kilowatt PWM pure sine wave inverters with large toroidal steel transformers, usually running from 48v for off grid home operation.
If you want to join the 48v club, Forum members have some very nice proven designs and vast experience and knowledge to share with you.
I don't think you are going to be very happy with a 12v system. Nobody does that these days, except in moving vehicles. Edited 2020-12-15 17:22 by WarpspeedCheers, Tony.
InPhase Senior Member Joined: 15/12/2020 Location: United StatesPosts: 178
Posted: 12:52pm 15 Dec 2020
Copy link to clipboard
Print this post
Thanks Tony! In my time lurking here I've gathered that 48 V is popular among this group. And it makes sense electrically as well to make the DC input voltage as close to the AC output voltage as possible. I'm not opposed to it. But from a monetary stance, I can't make that kind of outlay right now. I actually have enough batteries to get 48 V, but they are different brands of different ages. When I go to a higher voltage, I'd prefer my batteries be all the same.
At any rate, I'm trying to understand what it would take to make what I described in the OP work, even if it is an exercise in futility. I'm highly experimental and don't like to move up before proving a concept with what I have on hand. So even knowing that 12 V is basically pissing up a rope, the same physics should still apply and help me grasp the concepts involved.
What is it that causes HF inverters to release their smoke during surge? Is it just the small HF transformers dying? Could that be alleviated by putting a lot of capacitance on the high voltage DC bus to take up the strain? Indulge me for a bit and let's forget about the 12 volt input for now. Assume you have sufficient input current capacity. You have a push-pull DC converter that can deliver, let's say, 15 amps at 170 volts DC comfortably. You also happen to have a nice 5 kVA toroidal transformer wound 2:1 with a center tapped secondary such that putting 120 volts into it gets you 120/240 volts out. I have such a transformer. It seems that all the same principles should apply whether the input is 48 V or 170 V, so long as the source can supply the necessary current long enough to start your load. Unless there's some sort of hidden details of inverter physics I'm not keen on?
This is all largely theoretical, but I can't say I won't bash this up to satisfy my curiosity.
Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406
Posted: 01:58am 16 Dec 2020
Copy link to clipboard
Print this post
HF inverters invariably have small ferrite cored transformers. Anything that runs at 60Hz requires a much larger steel core, with a lot more copper mass.
Now what sets the final power rating on any transformer is always temperature rise. Temperature rise goes up at the rate of current squared. That very quickly sets the limit on power rating.
Now you can build a high frequency transformer that is really small, but it has very low thermal mass. A x4 overload puts x16 heat into it, and its going to last seconds before it bursts into flames.
Now a great lump of a steel core might weigh eighty pounds, and half of that will be copper. The heat from a x4 overload is going to be very easily absorbed not for seconds but many minutes, and do it easily.
Capacitors simply do not have the energy storage capacity to help with this problem. In theory ultracpacitors might work, but enough to be useful would cost thousands of dollars. Its just not a practical solution.
I might suggest you build one of the Forum inverters, but instead of using a single primary winding, wind it with four 12v primaries. Connect them up in parallel for 12v operation. Connect them up in series for 48v operation. Maybe series/parallel for 24v operation.Cheers, Tony.
Revlac Guru Joined: 31/12/2016 Location: AustraliaPosts: 1024
Posted: 02:48am 16 Dec 2020
Copy link to clipboard
Print this post
I have had the 48v 4kw HF inverters running, one on the house and one in the shed, no surge blowups YET....Maybe one day The 2 LF inverters built with help on this forum are running well and will be replacing the HF inverter in the shed mostly because of the surge requirements I have starting large motors around 150amp at 48v, another reason to go 48V instead of 12v or 24v. The HF inverter have survived up to 100amp surge motor starts so far, but one day it wont.
The LF inverter will now do the dirty work as some of these large motors have enormous starting current, As Tony has said LF inverter can handle the surges with more headroom than the HF, also for me I have half the idle current than the HF inverter and I'm happy with that.
Easier to repair something you built yourself. Edited 2020-12-16 14:03 by RevlacCheers Aaron Off The Grid
InPhase Senior Member Joined: 15/12/2020 Location: United StatesPosts: 178
Posted: 05:40pm 16 Dec 2020
Copy link to clipboard
Print this post
This is what I was looking for. These high frequency cores just don't have the thermal properties to cope.
That would be a doozie I'm sure. Quite the spectacle with all the primary windings.
I am working my way toward one of the inverters I've read about on here. I just have to get through the threads first. It's a personality disorder
Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406
Posted: 07:03pm 16 Dec 2020
Copy link to clipboard
Print this post
Not really.
Depending on the size of the core, something like four separate eight turn windings may be about right. Having that, instead of a single thirty two turn primary is not such a big deal. The same heavy gauge wire would be used in either case.
Just fit some big lugs onto the tails and bolt them together as required. Edited 2020-12-17 05:10 by WarpspeedCheers, Tony.
InPhase Senior Member Joined: 15/12/2020 Location: United StatesPosts: 178
Posted: 10:29pm 16 Dec 2020
Copy link to clipboard
Print this post
I wouldn't go through the trouble at 12 V, not for the inverter I intend to make. I would be happy at 3 or 4 kW and 300+ amps, even for short periods, is just more than I want to build for. The DC push-pull draws a little more than 250 amps when I max it out. I try to baby my batteries. Maybe if I had more parallel sets I would go for it. But if I had more batteries to make parallel sets, I'd just go with a higher voltage anyway.
I think I will go with 24 volts with the option to reconfigure to 48 volts as I grow. Is there a current write up about the various inverter topologies you guys have created here? I have been using the search function to poke around. So far, I gather there's the Is inverter, the Warpverter, the Nanoverter, and the Madness Inverter? The info is all over the place. Edited 2020-12-17 08:31 by InPhase
noneyabussiness Guru Joined: 31/07/2017 Location: AustraliaPosts: 512
Posted: 11:40pm 16 Dec 2020
Copy link to clipboard
Print this post
* ozinverter, was the start one.. the others excluding madness's are " evolutions " of it... originally the PJ , then the eg8010, but still basically the same concept. Low frequency, big transformer, bulletproof inverter.. HF have there place, low weight small footprint etc, but as mentioned you are limited by surge current etc. but if you are needing it for a particular load, it will generally be fine as long as you are generous with power ratings...
As example, i had a " 2kw " HF job running my pool pump for a short time, had a rock jam the pump and overloaded the inverter, blew it to smithereens ( and it had " overload protection " the same as my low frequency inverter, obviously wasn't fast enough). Yet my low frequency inverter had a direct short on 2 occasions and the most it does is trip the breaker... so topology plays a very big role ... just depends what your intended purpose is.I think it works !!
InPhase Senior Member Joined: 15/12/2020 Location: United StatesPosts: 178
Posted: 12:28am 17 Dec 2020
Copy link to clipboard
Print this post
I typed Oz, but stupid phone "corrected" me unnoticed
I have a 1000 watt HF PSW that won't start a 1/2 horsepower pump, UNLESS I start a bench grinder first. And the grinder has to have a hand start else the inverter won't even start it either. The bench grinder acts as a generator and feeds the pump motor what the inverter lacks. I'm surprised it works at all and I'm afraid it probably won't last.
Since my idea involving the DC converter will only lead to heartache and depression, I reckon it's time to get moving on Plan B. I've built a small PSW with the EGS002, but HF. Now I'll have a shot at LF. I don't particularly like the idea of being dependant on the Chinese for sine wave control. I'd like an alternative but don't have the coding skills to write a program to do it either.
Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406
Posted: 01:25am 17 Dec 2020
Copy link to clipboard
Print this post
All but the Warpverter use the exact same topology. A big bunch of mosfets, pulse width modulated at around 23Khz, driving a honking great big torodal transformer. All the variations of that involve how the pwm is generated and some various nice extra software features.
That is how I would do it myself up to about 5Kw. Above that it becomes increasingly more difficult to avoid problems. Not impossible, but you really need to know what you are doing. Basically high frequency pwm does not scale up very well.
If you want 2Kw to 4Kw any of the well proven Forum designs will work very well for you.
The Warpverter is an entirely different topology, it requires four square wave inverters which require four transformers and a lot of extra parts. Its just not worth the trouble unless you require ether extremely high power, or you wish to run it from an unusually high dc input voltage.
It scales up wonderfully well. There is really no upper power limit, and they are easy to build and get going because all of the switching is done at very low frequencies with just simple true square waves. Its a true sine wave inverter with distortion less than 1%. Anything over 5Kw, a Warpverter starts to becomes much more attractive.Cheers, Tony.
Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.
Print this post
