Grid Tie Inverter Schematic

[tcmtech]

To get a thousand dollar cost avoidance with 1KW back feeding continuously I would need to run it non stop for....

1000 / .12 = 8333 hours.
Or 347 days 5 hours.

or about 4 times that factoring in usable year round day light is maybe 8 - 10 hours a day plus the numerous cloudy days too.

Toss in 1KW of usable solar panel power. Thats what about $5000 more.
Now I am at...

6000 / .12 = 50000 hours
multiply by 4 for realistic run time.
200,000 hours or
22 years, 297 days, 20 hours
Maybe add one more year of runtime for parts outlay from near 23 years of general maintenance too if your dam lucky.
SO lets say 24 years average assuming my electric rate doesn't jump again in 24 years. (not holding my hopes to high on that one either)

And hope like hell hail, wind, snow, ice, design flaws or age doesn't take them down first!
Sorry but not cost effective in my mind.

 

[Val Gretchev]

Re: Line Powered PWM Rev B

Here is one for tcmtech. This one uses the International Rectifier driver chip IR2113. I don't have any experience with this chip, so let me know if I wired it wrong. Also, remember that this design is not completely isolated from the PV array.

 

[colin55]

The Australian government subsidy of $8,000 for a 1kW solar system is a total scam but isn't everything associated with a government a SCAM?
The effectiveness of the system is no more than 4-5kWh per day and any power fed back to the system is paid at a rate of 44cents to 60cents per kWhr. This is also a scam as it costs 3.5cents to produce electricity and householders pay between 10cents and 16cents per kWhr.
No wonder households are signing up for the scheme in droves and most companies offering the package are already 9 months over-booked.
Who would buy an inverter or panel when they all come FREE!

 

[tcmtech]

From what I can see your schematic of the power handling system is identical to what I use. As you can see you get full control of the entire H- bridge and full shut down with only 3 logic level control lines. I just use the 500 volt IC's because I typically dont need higher voltage but for a higher voltage setup I would not hesitate to use the 600 volt ones you have.

With this power handling circuit architecture its simple to make countless sizes in this design but still have only one master control board design that can operate any of them with only a few simple reference signal changes!
I smell dirt cheap mass production! Dont you?

 

[tcmtech]

So Collin let me get this strait. You sign up for this $8000 solar panel and inverter system subsidy and you get everything free?

Then when its hooked up and running you sell 3.5 cent a KWh electricity back to the utility that then pays you 44 - 60 cents a KWh for it likely with a independent meter system?

And then you use your regular house meter and buy it back from the utility for 10 -16 cents?

Can I skip the solar panels and inverters and just get the sell back meter?
Buy my power at 16 cents a KWH and then bump the voltage up a bit so that it will flow back through the sell back meter and I get paid 44 cents a KWH?

I could make far more pocket money that way than if I used the solar panels and inverters!

Something doesn't seem right here. I need a further explanation!

 

[Val Gretchev]

To: tcmtech

If only I could find a vendor that sells a choke and transformer at a reasonable price, it would be a design worth going all the way with.

 

[tcmtech]

What size choke and transformer do you need?
I may know of donor machines or devices to look for or may have a supplier that you could get stuff from if I know the size.
What cost do you consider reasonable?

 

[be80be]

I say I have any thing you need in my basement

choke and transformer

easy to make your own.

 

[Val Gretchev]

Re: Parts Cost

I don't care what the cost is for a single quantity prototype build. I could easily buy one from Digi-Key as long as they have it in stock. They list a bunch of transformers but they have 0 stock. The same is true of their high current reactor chokes. I figure I would need a 5mH to 10mH inductor but it has to carry the full current of the output which is around 10A for a 1kW inverter. The transformer doesn't have to be a high quality medical toroid. Any isolation transformer will do that has the right primary and secondary windings.

The problem is that, when you want to purchase these parts for production, the cost has to be much lower. The transformer has to be in the $100 range and the inductor should be around $40. There is no point making a prototype, in my opinion, if you can't reproduce it at a reasonable cost in any quantity.

A prototype raw (unstuffed) PCB would be about $500 for a quantity of 4 boards at PCB Express. In larger quantities, I could get the boards for about $30 each. However, doing a board layout properly takes about 40 hours of time. I wouldn't want to invest that kind of time ($100/hour) unless it was worth my while. However, sometimes I do a board like this just for the fun of it. Perhaps this is one of those times.

 

[Hero999]

There is no point making a prototype, in my opinion, if you can't reproduce it at a reasonable cost in any quantity.

That depends on why you're making the prototype. If it's just for educational purposes and you're not bothered about getting it made commercially then the single unit cost it more important than the production cost.

 

[Val Gretchev]

Re: Making Transformers

While we are on the subject of transformers, here is a video of how you make one. **broken link removed**.

While we are at it, if you were making a custom output transformer (T2 in the schematic), why not add an extra winding for 30 Volts center tapped at about 1.5 Amps to use for the low voltage power supply and eliminate the other transformer (T1) completely. That's another saving.

 

[Val Gretchev]

Solar Panel Optimizing

National Semiconductor has entered the end-user alternative energy marketplace with their SolarMagic(TM) PV Power Optimizer. You can read all about it at their site **broken link removed**.

At first glance, it appears that the SolarMagic(TM) technology is nothing more then a voltage booster with a maximum power searching algorithm. Fortunately for all you PIC-lovers out there, Microchip has already published an Application Note and source code for just such a device. Download the article and code at **broken link removed**.

This same algorithm can be applied to the main Grid-Tie inverter to maximize its power output. Start with the attenuator at near zero. Measure the PV array voltage and current and multiply the two values together to get power output. Increase the attenuator voltage. Re-calculate the power output. If the power has gone up, keep increasing the attenuator voltage until there is a power drop. Then back up the attenuator voltage one step. You have found the optimum operating point for the current illumination. Continue with the dithering algorithm to maintain the optimum PV power utilization.

 

[colin55]

National Semiconductor has entered the end-user alternative energy marketplace with their SolarMagic(TM) PV Power Optimizer. You can read all about it at their site SolarMagic? Technology Home -- Solar for the Real World.

At first glance, it appears that the SolarMagic(TM) technology is nothing more then a voltage booster with a maximum power searching algorithm.

It appears that "SolarMagic" is nothing more than a blocking diode - that's all you need.

 

[Mr RB]

National Semiconductor has entered the end-user alternative energy marketplace with their SolarMagic(TM) PV Power Optimizer. You can read all about it at their site **broken link removed**.

At first glance, it appears that the SolarMagic(TM) technology is nothing more then a voltage booster with a maximum power searching algorithm. Fortunately for all you PIC-lovers out there, Microchip has already published an Application Note and source code for just such a device. Download the article and code at **broken link removed**.
...

I designed a commercial solar MPPT a few years back with VERY high efficiencies (no I didn't use Microchip's system). I can't say too much about if due to contractual reasons, but I can say is that efficiency wise it is right up there with the best performing units in the world.

It's not that hard to do the max power point tracking itself, what is difficult is to do the actual voltage conversion reliably at 95+% efficiency and do it cost effectively so the device is actually worth buying and adding to a solar installation. Dollar cost effectiveness is everything.

As for the state of the industry I fully expect in a few years that any solar panel costing more than a couple hundred $$ will have a tiny switchmode MPPT converter built in.

 

[tcmtech]

Can you whisper some secrets in our collective ear as to how you design a 95% efficient switcher? Or even only 94% efficient as so not to break any contractual blah blah blah?

I have read up on some methods of using two or more inductors or transformers being ran with a switching time phase difference between them. Does that actually increase efficiency greatly or just power handling capacity?
Sort of a single phase Vs two or three phase type configuration.
I work with commercial three phase stuff all the time but have never played with its concept in a switching power supply application.

 

[Mr RB]

Well there's the obvious synchronous rectification, using controlled FETs as diodes, to reduce the diode losses.

What I will say is that inductors are everything. Besides being hand wound, we had to get the main inductor core material made overseas to my specification of minerals using some exotic minerals (basically aerospace quality inductor core material) to get the switching freq low enough and maintain the required magnetic properties. That might give you some idea! I'm not an expert on the magnetic minerals, and I don't think I've used magnetic formulae since 1983 in DC-machines class, I just did a lot of trial and error testing on the CRO with sample cores and refined it down to the right properties to suit our needs, then had a production batch of cores produced to that spec.

90% is pretty easy. 95% gets harder, 95% with good reliability, high current density, smallish size, large operating range and decent cost effectiveness gets real hard.

As for a multiphase system I have never done one, give me a nice simple buck regulator or push-pull voltage transformer anyday.

 

[shungun]

Does anybody know how to make an inverter from 12volt dc to 12 volt AC? around 6 amps.
i tried a squarewave generator with 2n3055 but didnt work too well.

 

[Val Gretchev]

Re: PWM Circuit

I built a test jig to verify the operation of the circuits shown in Post7.zip. I bypassed the attenuator chip and fed the AC signal from the voltage divider R5/R6 straight into the comparator. I also changed the value of C6 to 0.022uF to produce a higher frequency triangle wave. I put a small filter on one of the U8 outputs consisting of 15K ohms and 0.1 uF and took these pictures.

As you can see, the wave shape isn't exactly as the input waveform. Also, there is a considerable phase shift in the output which will be a problem. If anyone has any ideas as to how to eliminate this phase shift, please let me know.

 

[Mr RB]

Well looking at the phase lag, smoothing and reduction in amplitude that are obvious in picture DSC_2347 it looks like you are overfiltering something?

 

[Val Gretchev]

Re: PWM Circuit

I didn't have a MAX5437 chip on hand so I used a manual potentiometer and a TL082 op-amp. I wired the op-amp as a phase shifter and was able to bring the mains AC waveform and the filtered output in phase. I am now fairly confident that this circuit will produce the desired results. The next step will be to add the IGBTs and the output transformer.