diy solar

diy solar

DIY MPPT SCC?

Warpspeed, how well does your AC output over-current work ? I would imagine that output short circuit is the easiest to mitigate ?
 
Over current ?
We don't need to worry about minor issues like that at the inverter.
Just an ordinary 20 amp C curve thermal/magnetic breaker in the 230v output.
And a 63 amp dc breaker on the incoming 100v dc, with a thirty ohm soft start up resistor and a push button across the breaker.
The ac breaker always trips first, fault current mostly comes from 36,000uF worth of electrolytics.

Popped plenty of breakers over the last four plus years, low frequency square wave transformer inverters are indestructible.
I was nervous at first, but after four years, I now just treat it like the grid.
 
Over current ?

Popped plenty of breakers over the last four plus years, low frequency square wave transformer inverters are indestructible.
I was nervous at first, but after four years, I now just treat it like the grid.

Oh ? I thought yours was a sinewave inverter ?

The older Trace mod-square wave inverters were push-pull designs with a clamp winding and clamp FETs.

Then the DR mod-square was full-bridge. Can't remember what the current limit was ? Software I think.
FETs weren't as strong way back then though. Full die in the package helped a lot.
 
It is a sine wave inverter with less than 1% measured THD.
But it actually has four separate square wave inverters in it, that are added together to make up the final sine wave.
Its a 5Kw digital to analog converter !

That is what makes the Warpverter so unique.
The four square wave inverter secondaries are all connected in series to create a sine wave with 81 steps peak to peak.
The secondaries have different voltages, and are switched in and out, and when added together create a quite respectable sine wave.

These pictures are taken direct across each inverter secondary winding.
The oscilloscope is externally triggered and the vertical sensitivity has not been changed.
 

Attachments

  • Inverter1.jpg
    Inverter1.jpg
    462.9 KB · Views: 6
  • Inverter#2.jpg
    Inverter#2.jpg
    471.3 KB · Views: 6
  • Inverter#3.jpg
    Inverter#3.jpg
    474.7 KB · Views: 6
  • Inverter#4.jpg
    Inverter#4.jpg
    485.6 KB · Views: 6
Here is what happens when you connect the above secondary windings in series.
Magic happens !!

Each separate inverter is just a dumb square wave inverter.
But working together 0.85% unfiltered distortion.
 

Attachments

  • Inverter#1 + #2.jpg
    Inverter#1 + #2.jpg
    490.3 KB · Views: 4
  • Inverter#1 + #2 + #3.jpg
    Inverter#1 + #2 + #3.jpg
    496.4 KB · Views: 4
  • Inverter#1 +#2 +#3 +#4.jpg
    Inverter#1 +#2 +#3 +#4.jpg
    492.1 KB · Views: 4
Ok, it is like the old Trace Engineering SW5548 OR SW4024 but they only had 3 sections on series, each consecutive section smaller. It's THD wasn't as low as yours but still only a couple %....

It was also grid tied and charger because it is bidirectional, like yours.

Different waveforms for regulation were pulled out of a waveform ROM

Are you using M6 laminations? The thinner, lower loss type?
 

Attachments

  • Copy of SW-Waveforms.gif
    Copy of SW-Waveforms.gif
    268.9 KB · Views: 5
Last edited:
Same principle as the Trace, I put together the first one of these just over forty years ago. Even back then it was not a completely original idea.
At one time, I had a transformer manufacturing business, but that was very long ago.
I still have a buddy nearby that is in that game, and he wound four transformers to my specification for just the cost of the materials.
Yes, the thinner grain oriented silicon steel laminations, and run at a low flux density. Idling power is just over 30 watts for the whole inverter.

Toroids would have been very slightly better, but I am in my seventies and cannot be stuffed doing all that myself these days. My buddy has all the gear to dip and bake the transformers and they came out looking like professionally wound transformers, which they are.
Every other Warpverter uses hand wound toroids, and they all work very well.
A split phase Warpverter for 110/220v would be entirely possible, but I am not aware of it ever being done.
 

Attachments

  • 2018-10-10_0001.jpg
    2018-10-10_0001.jpg
    989 KB · Views: 11
Last edited:
Your transformers are awesome ! At Trace, we made our own power transformers. It was handy to be able to make a change and have a transformer to test the next day.

Somewhere I have a copy of a chapter of a very old book about inverters from around 1960 I got from the library. Chapter was called something like "Improving the output waveform".

It used this same basic technique but it used thyristors instead of transistors. SCRs or TRIACs.

I still employ TRIACS or SCRs when it makes sense. Usually triacs.

boB
 
I find it particularly interesting that you were at Trace Bob.

The transformer design is actually pretty interesting, the current in all of the transformer windings is a constant continuous unbroken 50/60Hz sine wave, but the voltages across the individual windings are all rectangular waves containing various switching frequencies.
The functioning of this is conceptually, a pretty difficult thing to get ones head around.
It took me quite some time to figure it all out and finally get really good results from the transformers.

I discovered some short cuts to vastly simplify transformer design, and how to do it is fully documented over at The Back Shed.
Many successful Warpverters are now running all around the world. All are in the 5kW to 8kW class.
Its all free open source, I have no commercial interest in it. Its far too expensive to build commercially, as Trace will know.
But its PERFECT for a home brew project using recycled materials to wind the transformers.

Generating the mathematically correct drive waveforms for all the ROM lookup tables was pretty challenging too. Software is not something I excel at. Several of the professional software gurus over at The Back Shed individually solved that problem embarrassingly quickly.
For me originally, it was a very long drawn out process indeed.
Its all been a fascinating development project over the years and very gratifying.
 
Last edited:
I find it particularly interesting that you were at Trace Bob.

Yes, here is a very quick history...

I started at Trace in 1994 after working in pro and consumer audio and a few years in medical electronics.... Electronics part of those.
Brother Robin started at Trace in 1990 after designing audio gear and then military flight power supply system stuff for several years. Mechanical design part of the electronics mostly.

We were at Trace until and slightly after Xantrex and Trace merged. . XanTrace (as I called them) bought Heart Interface who also had local roots, including people from Trace and were in Kent, in the south Seattle area. So we also had some interaction with them.
Xantrex, the Canadian company took all that over though which forced us to start OutBack Power

After 5 years of OutBack and a falling out with our "partners", Robin started MidNite Solar. I worked at Magnum Energy for 3 and 1/2 years in engineering until OutBack bought my stock and then it was MidNite Solar all the way. The story isn't nearly that straight forward but that's the basics.

All of those companies started with a local high power audio amplifier company called Phase Linear (Bob Carver) in the very early 1970s in Edmonds Washington where most of us worked at least for a while. PL made their own power transformers so that is where that expertise started. Bob Carver was basically the "seed" for MANY different companies around the area... Audio and inverter companies, many of which I either worked at or was associated with. EXCEPT for Dynamote which was another inverter company from the mid 1980s in the Seattle area was eventually sold to another company which continued selling to service vehicle/truck companies like for aid cars, fire engines, etc. It was a popular company for a while and eventually did grid tie. I can't remember the name right now.

Two of the Dynamote engineers came to Trace and one went to Magnum. We were pretty much all friends and worked with each other at various times.

All of these companies manufactured up here in the northwest in Washington until 2010 to 2015 or so. And except for MidNite Solar, all are manufactured either in China, India or Mexico. Exeltech I believe are still made in Texas ? We still do our own SMT and through hole and inductors, etc. The transformers for the newer HF inverters are a bit harder to do economically on a production basis and we may require some overseas help with those.

I have an older video that Robin, I and our dad did of Trace Engineering, including the transformer area you might be interested in seeing.
It is a copy from VHS tape and is a bit faded but you can still see some interesting things.... No sound. Transformers start around 3:50



Generating the mathematically correct drive waveforms for all the ROM lookup tables was pretty challenging too. Software is not something I excel at. Several of the professional software gurus over at The Back Shed individually solved that problem embarrassingly quickly.
For me originally, it was a very long drawn out process indeed.
Its all been a fascinating development project over the years and very gratifying.

Yes, I know ! It is quite a process.
As I remember, Greg (SW engineer) had a pascal program that came up with the different ROM tables for the SW.

It has been a great ride working with all the Seattle area electronics companies. I have learned a lot.
Getting old sucks but I'm not quite ready to stop learning yet !

boB
 
Last edited:
Ah ! I remember Phase Linear and Bob Carver very well.
Still have a series two Phase Liner 400 amplifier here right now in fact.
Still an excellent amplifier.

That film clip is fascinating, I have never seen a Trace product before, only ever heard about them, and never really knew much about them or any of the details. There must be very few if any made it here to Australia.
The three transformers and the stepped waveform all seem very familiar to me though.

All my earlier designs only used three transformers, and it was perfectly adequate.
Its only in more recent times I experimented with four transformers, and it makes a huge difference to THD.
My rather unusual feedforward voltage regulation idea eliminated a lot of parts too, and improved performance even more.
Mine does not do grid tie though.

I also found it interesting that the voltage feedforward, and the Hall feedforward current regulation worked equally well when feeding power back into the battery when a grid tie inverter is coupled up to a Warpverter for full off grid operation.

Another way to build a simple sine wave inverter you don't hear much about these days, is the use of a ferroresonant transformer. I worked for a while as a designer at Sola Basic, and designed one of their extensive range of ferroresonat UPS's.
All fascinating stuff.
 
Does anyone have a copy of Trace SWCA software version 1.01 or1.02 ? I can't find my old floppy copy.
Any help would be greatly appreciated.
Thanks ,Ole Dog;
 
Back
Top