New no battery pure sine wave solar power system with smart switchers.

[GridWorks Green Solar]

Good morning
Thinking about how to command the secondary power board from the primary CPU, the center four FETs is all that is needed to produce pure sine power, the outermost four FETs are directly tied to the inner four so you can easily double the output, in the previous attempt to sync I used the four FET trigger signals and a the CPU trace marked (gnd) (total 5 wires) on the CPU to copy the signal to the other power board.
For what ever reason the first attempt did not work, in the early days of computing we would double the RAM on our computers by stacking the chips on top of one another and soldering all but one pin that went to another controller chip.
I think my next attempt will be (12 wires) to duplicate the center four FETs to the secondary board (3 wires for each FET) may also need to run the CPU (gnd) wire to the secondary board.
Wish me luck ? I really need this to work.?

 

[halfagascan]

you need to review, back to out of focus pics.
I would like to follow along, but, honestly, you ramble and repeat

 

[GridWorks Green Solar]

you need to review, back to out of focus pics.
I would like to follow along, but, honestly, you ramble and repeat

I normally take six to eight pictures of any given item and review all the pictures deleting everything but the best picture and upload it, most images on my phone are around 2 megapixels but when looking at the size of the pictures online they are more than half the original size.

Thanks for telling me and if a drawing or picture you need is unreadable please specify and I will do my best to redraw or rephotographed the item I will be happy to do a close up for example the dual DC input module or anything you need, parts used,parts ratings, meter readings I mean anything to help you understand or build one of these amazing power systems.

Because of a mistake in testing I assumed the second ISO transformer was not needed and that sent me into a endless loop of bias issues that caused confusion and expensive parts failures.

I have only solve the issues in the past week or so and just when I knew I reached the point of success in the seamless DC dual power input module, managed to surge the 2000 watt board and damage the FETs again, until I did the autopsy I really thought it was another bias issue adding to the confusion.

Sincerely sorry about that, also rushing to get in a full day of testing before going out of town for the Holiday.

What is important here I have a working system that transfers power seamlessly between solar and the 120 VAC secondary input of your choice, and because the PV system 180VDC directly drives the pure sine power output board and completely bypasses the normal boost section found in most inverters it may very well be the most efficient way to do solar power period.

The system is small and compact except for the normal size solar panels and the one or two ISO transformers needed to block the bias of grounded power systems.

The system is already expandable 1000/2000/3000 watts but I hope in the next couple of weeks to link power boards at the CPU level so you can expand easily without replacing the power board you already own, I have no plans to go past 3000 watts, but in theory with synced power boards their is no limits to how big of a power system you could build.

Heading back home tomorrow, expect in the next couple weeks to expand to 3000/2500 watts pure sine, It will take nearly all my PV panels to test this configuration but I already know that it will work.

Have a good evening, I have an eight hours drive to get back home tomorrow.?

 

[GridWorks Green Solar]

Hi all, made it home safe, late afternoon 4PM on a heavy overcast day, replaced PV neon indicator and normal restart of the seamless power system.?

Tomorrow will start testing to sync multiple power boards.

BTW better diagrams and help documents soon just need a couple more successful test to lock everything down once and for all, at this point I know it works good with careful full loading and real world side by side fridge testing I am going to be certain how best to use this solar power unit.?

Good evening ?

 

[GridWorks Green Solar]

At the risk of rambling on, I am going to bring up the first attempt to charge the EV car and how the FETs melted on the 2000 watt board.
The system was configured for 1700 watts on a 2000 watt board, should not have attempted the test without at least 1950 watts of solar PV power but this is what really happened.
The car did charge for a full 10 seconds before the power board failed by 4 of the 8 FETs internally melting, first off no way should have 1700 watts of PV power damaged that 2000 watt capable board, the fact it ran 10 seconds and melted FETs speaks to the brute force kick of these new power systems can deliver in full sun. Before the car charger test that same configuration ran the house lights and fans and a dual tube quartz heater for a full load test.

Will need to proceed slowly with this type of testing but early results point to something extrodenary coming in the next few weeks.?

 

[eemarty]

I'm also trying to follow along here. Could you tell us what are the models of "CPU" board and "power" boards you are using and where are you ordering them from?

Do you have a clamp meter that can measure inrush peak values? It might help to see what the peak current draw is for your car charger.

 

[GridWorks Green Solar]

I'm also trying to follow along here. Could you tell us what are the models of "CPU" board and "power" boards you are using and where are you ordering them from?

Do you have a clamp meter that can measure inrush peak values? It might help to see what the peak current draw is for your car charger.

All parts come from China with a two or three weeks delay, first discovered them on eBay after searching for replacement pure sine power boards, their are a couple of sellers that always seem to have them 1000/2000/3000 watt modules checked as of yesterday.

I don't want to promote anyone or anything I just use things that work, after opening up many inverters have noticed that all the output power boards are slightly different so their must be several companies making them. The EG8010 CPU driver board is preferred because of the perfect sine wave power it creates, standard level one EV car charging is 1400 to 1950 watts of 120 volt AC, my clamp on amp meter is AWOL but I can hook up a current sensor and provide you with exact numbers soon.
Off to a great third day of testing the seamless power system.?

2 PM crappy solar day, switched on all the lights and took this picture.

 

[GridWorks Green Solar]

Here are the exact readings from level 1 car charging for my KIA soul EV from grid, the power ramped up over three seconds period to 1270 watts 11.05 amps 114.9 VAC.

By the numbers the solar EV charging test should have worked.

Next time I hope to have a more robust unit for car charging.?

 

[eemarty]

So are you sure that 11.05 amps is the peak and there aren't' short bursts that are higher?

 

[GridWorks Green Solar]

Watching it also on the "O" scope looking pretty smooth to me.

I will have to use this same setup next time to monitor solar charging, something missing in the puzzle, scope can record realtime data.

Great third full solar day with the seamless power system, not going to load up/full load the system till I get the 2000 plus watt combo system working.
The sky was mostly overcast but the sun made a few pop through events that were good for data collection, this system is going to run lots of your stuff my big question right now is how well will it start a refrigerator compressor?

More to come, mainly today just preparing for making magic tomorrow, going to test with 18 VDC until I see the output on the dual boards.

 

[eemarty]

I'm still just trying to understand what you are using. So what you are calling the "CPU" board is one of these right?

https://www.aliexpress.us/item/2251832454994971.html

Could you please send a link to the "output power board" you are using?

 

[GridWorks Green Solar]

Yes, you found the CPU driver board at a great price, here is an example of a fully assembled 1000 watt board that already has the CPU board mounted to it.

NEW 1000W Pure Sine Wave Inverter Power Board Post Wave Amplifier Board Finished | eBay

Find many great new & used options and get the best deals for NEW 1000W Pure Sine Wave Inverter Power Board Post Wave Amplifier Board Finished at the best online prices at eBay! Free shipping for many products!

www.ebay.com

I currently need two of the CPUs to replace damaged ones, but todays full focus is making the CPUs I still have working sync multiple power boards.?

If I fail probably going to order three of the small CPU boards.

Some confusion in how they size the units, my recommendation is using two of the 1000 watt kits to make a 2000 watt board you need 8 FETs on the power board to be at 2000 watts level, have noticed different part numbers on the FETs maybe more capable but as a tech I know it is better to spread the heat load across eight spots instead of just four, the 3000 watt board looks totally different and I have never actually owned one.

The boards are actually pretty tough the 1000 watt board running the house lights, fans, security cams is about 5 years old, use the ISOLATION transformers to keep them safe from any bias from grounded or negative ground power systems.
Always turn on the CPU last in the start up sequence and for soft restarts of the system the CPU should be feeding off the secondary power input this is all about giving a fully loaded system a soft start vs a big surge that can likely smoke the FETs.

Good luck ?

 

[GridWorks Green Solar]

Way early to claim any sort of victory taking a break right now but at 18VDC logic probe and "O" scope says I have done it.
Repaired and synced a 2000 watt to a 1000 watt board, it will take many steps to build them up to handle the 3000 watts case,heat sinks and more.

Only took four wires and removal of the bad CPU from the 2000 watts board, also keeping all the pluses and minuses correct.?

Calling it a night increasing voltage did not work as expected, and I don't need any more stress today.

Good news 1000 watt board still working after cutting jumpers and disconnected 2000 watt board, next test will be with shielded coax cables for the four triggers.
See you soon.

 

[GridWorks Green Solar]

Good morning,

Finally just a bit of success in syncing power boards.
This is what I have learned in about six hours of bench testing.
1. The CPU driver board can only switch 8 FETs maximum.(easy to combo two 1000 watt boards)
2. Only takes four jumper wires on the FET triggers and a shared negative (-) on the high voltage side to sync the power boards.
3. The secondary board needs the CPU removed and no CPU power is required for that board.
4. Attempting to drive more than 8 FETs won't work but it also won't damage the primary CPU.
(Strangely at 18VDC input it can switch 12 FETs and I have the "O" scope sine wave outputs to prove it)

Much more testing required as to combining power board outputs for double watts.
First test of combining the 120 VAC outputs from the individual power boards is a total success, claiming victory.?

Still need to order a new CPU for the 8 FET 2000 watt power board.

Seriously angry that have tossed out in the trash six of the 1000 watt boards that could have been easily cleaned up and reused as secondary synced power boards.
The damaged low voltage center part isn't used at all in the secondary power board role.?

 

[GridWorks Green Solar]

The seamless power system has performed beyond expectations for 5 days now, as of tonight it will be shut down and upgraded to a combination synced 2000 watt unit, not the 3000 watt unit I wanted but will keep the project moving forward and allow for larger loads like a full size fridge and more.
The system is currently configured for 1700 watts the 2000 watt synced combo unit should allow for some impressive testing, expect to be down a couple days with updates and making darn sure the FETs are properly heat sinked and cooled.

In case you are interested, I will be adding a synced secondary 1000 watt board to the 5 years old primary 1000 watt board, this will require moving the diode bridge a few inches to make room in the case for the new board.
Both power boards have been featured in previous photos and testing, Absolutely want to see this unit in action ?

Ordered two new CPU driver boards, one is to get the badly needed 2000 watt 8 FET unit back in operation.?
Have already prepped the unit for the new CPU.

See you soon ?

 

[eemarty]

So If I have this correctly you are buying this board and modifying it to add the 4 additional power fets to make it 2000W and then powering it directly from a string of 10 series connected 12V nominal panels for about 190VDC to generate a 120VAC output from the board. Is that all correct?

Do you have schematics anywhere? I haven't found any documentation in English. I found this manual in Chinese, but I'm not sure if its the same one for the kit you ordered. You can zoom in on schematic 6.2 on Page 8. I believe that's the one this kit is based on. It looks like the kit also has a little regulated power supply to generate the +12VDC and the +5VDC from the 16VAC input which is not shown on the schematics I found.

How are you handling grounding? Are the negative of the solar panels and the negative input to these power boards both earth grounded? And you also use a high power isolation transformer on the output? How is that grounded?

 

[GridWorks Green Solar]

9 panels standard 12VDC 100 watt 180 VDC fuse with 250 volt 5 amp quick blow glass fuse. 180 volt is the magic no matter what PV array with multiple parallel 9 panels you can expect 800,1700,2500 watts of pure sine power, the 10 th panel is used in total off grid situations to drive the power board CPU and fan. The ground is the enemy of this system grounds are added and deleted at the Isolation transformers, Insulation and Isolation are key to making it work.
You can ground the case but you need to float the power board on rubber/silicone or similar unit comes with isolation heat pads for the FETs looks like four gray stamps.
Solar panels are not grounded and if you add a ground that pretty power board will explode, not kidding!

Unfortunately no schematics used a preassembled 1000watt board as a guide to build two 1000 watt kits into a 2000 watt board, easy as just adding the missing parts on the circuit board, just 4 additional FETs and the surrounding resistors and diodes, working the first time.

Follow the drawing/attached photo closely, ground is only added on the output of the ISO transformer only. This circuit diagram has been audited for safety and function but a ground bias will eat the diode bridge or power board if you break isolation rules. Diode bridge also needs heat transfer isolation pads and thermal paste.

Cool you keep finding things I need at lower prices, thanks.?

 

[GridWorks Green Solar]

Have a new shop bench to work from, no more dinning room table, have just started the 2000 watt synced combo unit.?

 

[eemarty]

9 panels standard 12VDC 100 watt 180 VDC fuse with 250 volt 5 amp quick blow glass fuse. 180 volt is the magic no matter what PV array with multiple parallel 9 panels you can expect 800,1700,2500 watts of pure sine power

You mean series not parallel right? With 9 panels at 100W each you wont get much more than 900W...

When you show a ground symbol on your drawing is that the earth ground rod for your house connected to your neutral in your main electrical panel?

I'm unclear why 2 isolation transformers are needed. Isolation transformers are hundreds of dollars each for a 1000W unit.

 

[GridWorks Green Solar]

Each string 9 panels in series 180 VDC each fused at 5 amps , first string will net 800 watts pure sine power.
If you parallel a second string of 9 panels you will net 1700 watts (still 180 VDC) from a 2000 watt capable power board.

If you parallel a third set of 9 panels you will net 2500 watts (still 180 VDC) from a 3000 watt capable board.

The reason the watts numbers run the way they do is the power board needs approximately 100 watts to produce the 122 VAC power, this debt is paid by the first set of PV panels making the 900 watts of panels produce approximately 800 net watts of power, 18 panels nets approximately 1700 watts,27 panels should net 2600 watts on a 3000 watts capable board.


Most people pay big for batteries that eventually go bad and need replacement, ISO transformers don't go bad and don't need maintenance.

Try and find old medical grade transformers on the cheap (They never go bad), If you only build a 800 watt system, two really nice New 1000 watt transformers on Amazon under $400 before shipping.

Sorry I went way out of the way to eliminate the second ISO transformer turns out that is exactly what was needed to fix multiple problems for the project.

Natures rules running this project not mine, absolutely amazing how things have developed in this project.?

You can go off grid mode and not use Isolation transformers but the thing gets touchy around grounded appliances and such. Connecting the CPU power board to grid without using the door bell isolation transformer will instantly explode the center of the power board. Honestly the transformers make things run safe and smooth, If you are going to daily use for solar power do things correctly and up to code.

Ground symbol is earth ground, for testing and use I use the ground from a grid powered outlet, In off grid situations putting down a proper ground rod for the ISO transformer is recommended.