Fieldlines.com: The Otherpower discussion board
Homebrewed Electricity => Controls => Topic started by: clockmanFRA on March 16, 2016, 04:36:33 AM
-
I will start a new Topic.
Those of you that have been following oztules wonderfull creations, may have noticed that the weak link for this OzInverter is the PCB Boards.
Don't get me wrong, there is nothing wrong with the 15kW Power Jack Boards, they do the job excellently!
The problem is getting them. Power Jack obviously get them in/made somewhere, but as I have put a book together, sort of 'Open source' concept, its embarrassing to find folk can not get them.
So with 'oztules' help I think I can move the Boards forward so folk can make or get them made.
I am at this moment putting some 10kW -15kW power boards together.
They are based on the W7 Powerstar schematics and the PJ boards.
Layout will be similar to the PJ 10kW - 15kW power board but modified, double sided, but the heat sinks may be different size or a stock size readily available, certainly not smaller.
Steep learning curve on making the main board, at 278mm/ 11inch long by 172mm/ 7inch wide, (just fits on a page of A4, that's handy).
I last did PCBs in the late 1980's, and Interestingly the manufacturing process for PCB's, 30 odd years later doesn't seem to have changed much..... :' ( Except the CNC milling machine approach. I thought about that, then remembered at this large size the blank boards tend to distort easily, so back to the photo etch process. :'(
These boards are large and I looked at getting some prototypes done/ outsourced, but the costs are silly, hundreds of $.
Also I note everyone wants Gerber files, and yet I scan in my present boards, alter accordingly on the PC, then the prototype folk say, "can only cope with Gerber files". :'(
I think the only thing I have left from the 1980's is a vertical etch tank, with aerator and heater, and some copper etch crystals. So best get busy getting everything else.
-
ClockmanFRA
What type files are you normally using. With all of today's open-source stuff, there's probably a way to transfer from one to the other.
I used to use Eagle format for PCBs
-
Remember, the gate resistors drop down to 5R6, and have a signal diode across them ( anode facing gate ) for the 8010 version.... big difference in heat.
................oztules
-
Hi ozules, okay got it.....
The new Control card, I am in your hands on that one, with the New 8010 chip/processor changes the arrangement I was planning.
So if I understand this correctly each of the 47r resistors at each gate, 24off, needs changing for a 5r6 0.6w metal film resistor. ?
And a signal diode, say IN4148 parallel to the 5r6 resistor with the anode facing the gate. ?
Here are the one piece Power Board, circa 2013, from Power Jack. This is the simple board I would like to update and make for our use?
The 0.6w metal film resistors at each FET are 47r and 20K. The exception is those 2off 2w 47r metal film resistors on each side row of FET's.
The only difference I could find between the Old PJ board and the New, circa 2015, board is obviously the 4 sub/daughter boards, but on those the resistors are still 47r and 20k, and the extra signal diode and the 24k and 22k with capacitor, but no 2w 47r resistors.
Both the new and the old PJ board has the small ferrite bead? 4mm long x 3.5mm dia x 1.5mm hole, on the centre leg of the FET, do we still need them?.
Thanks
-
Oztules, the old board had IRFB4310 FET's, (approx 18Meg) I would like to fit IRFB4110. is that Okay?
And the Old one piece PJ Power Board we hope to modify and re-build...........
Layouts for reference.......
-
Hi Bruce,
I am a clean sheet, I have NO PCB production software whatsoever any more, what I did have in the 80s was share wear etc, just won't run on my PC any more. Suggestions?
I would like to just amend and modify that old PJ Power Board and have it re-made, so I would like to keep things simple.
I just have a vast amount of jobs/work I need to do here.......
-
Eagle AFAIK is still free for the lite version. Pretty good though it has it's limits.
I took a quick look at their site and manufacturer's can even use it to create gerber files.
Hope that helps
Bruce S
-
We use Eagle for small boards. Max is (IIRC) 2 layers and about 8x10cm, for non-commercial work.
-
Yes, 47r replaced by 5r6, and diode in parallel with the 5r6 anode to gate... this for the egs002 setup.
The small board ...........I am currently is redesign mode to get rid of the propriety 002 board, and just use available chips.... through hole and plug in replaceable.....true self fixing in remote area design is what I'm aiming for.... rather than where folks are now frightened of their inverters. The 002 does this already, but I want to be beholden to no one.... I want chips only, no foreign boards. Plug in processor, drivers and we are good to go.
I note the internet problem in France, and will put a story on the little 002 inverter here on fieldlines when time permits, so you continental folk can see it more readily.
results so far on the test one are uber impressive... I just can't get it to blow up any more... ever since I took to it with a razor an hacked off the 393.
It has been running the house for a few days now, nothing bothers it, and we drive it hard............ bit like Chris Olsen does, sees 8kw spurts regularly, periods of hours at a time of 3kw and every where in between....... 14-16kwh/day at the moment........ very impressed with it... and thats with a substandard transformer... really only wound for testing at 1-3kw. There must be clipping somewhere in there, but I have not seen it when scoping it....
-
Thanks.
Yes, oztules, I have ordered some 8010 chips plus those holders, in anticipation......
Bruce, Damon, I looked at that software, sadly the board I am making is a big board, and to spend $600 on bigger boards production is just not cost effective.
I looked up Thermal Image and heat transfer on to the copper boards. It looks promising. But the printer must be a Laser toner image that can be heat treated using a laminator onto the PCB board and the paper peeled away leaving a mask image on the copper. Sounds good but will practise.
Crikey Laser mono printers have come down in price these days.! New about $60 for a branded make. And as I will have to add a few insert pages, the new laser printer can also do that. Pages will be erratum and direct scale PCB masks, for these boards. Folk can then photocopy at their laser printer copy shop onto the special heat transfer film/paper and etch their own boards..
This Power Board fits neatly on to a sheet of A4, so at this moment I have started using my CAD, I use it all the time and it has symbols, and is very very accurate and I am used to it.
There will be 3 mask for each side, here I am starting the Holes drill mask, just about to put the FET holes in.
whoops just noticed mistake, wrong year...... Its late evening here......
-
I looked up Thermal Image and heat transfer on to the copper boards. It looks promising. But the printer must be a Laser toner image that can be heat treated using a laminator onto the PCB board and the paper peeled away leaving a mask image on the copper. Sounds good but will practise.
Crikey Laser mono printers have come down in price these days.! New about $60 for a branded make. And as I will have to add a few insert pages, the new laser printer can also do that. Pages will be erratum and direct scale PCB masks, for these boards. Folk can then photocopy at their laser printer copy shop onto the special heat transfer film/paper and etch their own boards..
A few tips...
- Use inkjet matt colour printing paper, not more than 100g weight - check that the matt coating doesn't melt at laser fuser roller temperature
- Make sure the copper is really clean, iron on the laser printed image with the iron at 150-180C (you want 'hisss' not 'HISSS' temperature!!)
- Don't press too hard otherwise you'll squeeze the ink out sideways
- Allow to cool and then soak in tepid warm water with a bit of washing-up liquid in it
- The coating on the paper should separate from the fibres of the paper and stop furry edges and become fused into the thermal laser ink
- Clean off after etching with xylene (Hammerite thinners) or similar
Its a few years since I've used this method but I may have to go back to it as I no longer work at the place where we had prototype pcb etch facilities!
-
well for completeness, here are the component o/lay and pattern I ran for this experiment. I have made it as dead simple as absolutely possible... was just a test anyway, but it works as well as a top range inverter.... who would have believed it could be so simple??... it just works.
Day three running house.. no sign of glitch or stress no matter the load or complexity of the loads.
[attach=1]
[attach=2]
So put this layout into an image program, and resize until the 17 pins are exactly .1 inch apart... and print it out and etch it. I use a bubble jet printer... seems more reliable with black without fade areas.
I use normal A4 paper, kinsten pre sensitized boards, and olive oil ( or anything oil I have handy... even crc....) and a desk fluoro to 25 -30 mins
Print out the pattern on the A4 letter paper, place it on the kinsten board, add oil... goes transparent to UV light.... and 25 mins under the flouro at 2-3 inches or there abouts distance from pattern to light globe
I use NaOh for developing... carefully.... the I use CuSO4 and HCL for etchant...... the farmers use the CuSO4 ( in 25kg bags) for copper in their dams and fertiliser etc etc, and the HCl is available everywhere as brick and concrete etchant...1-25 liters containers., this is an island remember...
Yes the pattern could be improved immeasurably.... but for a quick test.... it worked well beyond my wildest dreams.
More info here: http://www.anotherpower.com/board/index.php/topic,1116.0.html
Will do a story for this site for the French folks too maybe....
..............oztules
-
My preference for a home brew board is to stay clear of smd's. Clockman it is a bit hard to workout a circuit diagram from just the photo's without having a PJ as a sample. Correct me if wrong. IS the main board basically a buss bar with the big caps and the daughter boards just hold the heat sink mos fets plus some resistors mini cap diode? A sketch of one leg would be great.
Futurlec can make pcbs in Thailand if there need for enough boards it may be worthwhile to do a run.
-
Yes make the 20k through hole.
-
Hi Frackers,
Thanks very much for those tips on the thermal image on to copper.
I was going to use Mrs laminator machine, but perhaps not! Now where did she put that Iron......
Hows your hard core Ozinverter coming along? For those of you that don't know his endeavours, he wound his own silicone iron cores, now that's dedication.........
-
well for completeness, here are the component o/lay and pattern I ran for this experiment. I have made it as dead simple as absolutely possible... was just a test anyway, but it works as well as a top range inverter.... who would have believed it could be so simple??... it just works.
Day three running house.. no sign of glitch or stress no matter the load or complexity of the loads.
So put this layout into an image program, and resize until the 17 pins are exactly .1 inch apart... and print it out and etch it. I use a bubble jet printer... seems more reliable with black without fade areas.
I use normal A4 paper, kinsten pre sensitized boards, and olive oil ( or anything oil I have handy... even crc....) and a desk fluoro to 25 -30 mins
Print out the pattern on the A4 letter paper, place it on the kinsten board, add oil... goes transparent to UV light.... and 25 mins under the flouro at 2-3 inches or there abouts distance from pattern to light globe
I use NaOh for developing... carefully.... the I use CuSO4 and HCL for etchant...... the farmers use the CuSO4 ( in 25kg bags) for copper in their dams and fertiliser etc etc, and the HCl is available everywhere as brick and concrete etchant...1-25 liters containers., this is an island remember...
Yes the pattern could be improved immeasurably.... but for a quick test.... it worked well beyond my wildest dreams.
More info here: http://www.anotherpower.com/board/index.php/topic,1116.0.html
Will do a story for this site for the French folks too maybe....
..............oztules
That's excellent oztules, nice layout, so you have a working Control card with that egs002 card with that 8010 chip.
I do love robustness and simplicity.
The small board ...........I am currently is redesign mode to get rid of the propriety 002 board, and just use available chips.... through hole and plug in replaceable.....true self fixing in remote area design is what I'm aiming for.... rather than where folks are now frightened of their inverters. The 002 does this already, but I want to be beholden to no one.... I want chips only, no foreign boards. Plug in processor, drivers and we are good to go.
.
Here here!,.......... or whatever means, I totally agree with you 'oztules'. And I say and repeat again, 'Thanks very much for your openness with your endeavours' its very refreshing.
-
Phred,
If you look through 'oztules' 60,000 odd words, you will see the W7 Powerstar Schematic, the PJ Power board is based on that. However, as you have already seen for the 'New 8010 OzControl Board' we need to change from 47r to a 5r6 resistor on all 24 fet's and parallel a signal diode, as the 8010 is a different purring beast to the PJ Control Board.
The Power board I am putting together is based on the PJ layout, as its robust, simple and okay for cooling installation. But there are several improvements I will make.
Obvoiusly, ....no SMD stuff to the board. .........I love through hole board components as they are good to work with as a DIY job. Soldering will have to be good. Where tracks go through the board I will use pins and solder each side. But fortunately there are not to many small tracks that require this.
I note that the PJ board has solder collars in most big holes to ensure good connections with some of the big connections, ie, 6mm hole for the 6 negative cables to the battery, FET legs etc. So if a connection is required on both sides with the component passing through then I will investigate further and amend.
My Mrs reckons I always over engineer, but with this Power Board it needs to be right?
Okay, As some of you will be aware, I have recently published a small A4 size, 76 PP, 151 colour photographs & 14 diagrams Booklett. "Making a Low cost, simple and robust 6kW, 50HZ Pure Sine Wave, 48vdc to 230vac, INVERTER." Its all in there, except these new Boards, which I will ad in as a pull out suplement with the first run of just 100copies.
Why did I do it, as there is no financial gain in these type of books.
1................ Up the Nose of the greedy ignorant Big Boys, especially SMA, its a long story!...
2............... With 'oztules' can do and will do direction, it is real and possible.
3............... Its 'Open source' type of publication. Like a famous Wind Turbine Man's publications, but remember I am just an apprentice constantly learning.
4............... Personal empowerment and self-improvement is very much our Philosophy here in Normandy, especially with real hands on cost effective renewable energy creation. Hopefully, I can pass on my experiences to other interested folk and importantly the next generation, especially my boys.
So in a few weeks time, when the bits come from across the globe, I will get the Mrs to put that extra page on our Web site for buying the book with everything in it, and the components lists & masks for making your own PCB's.
There is very little out there about the actual making stuff, and I like to believe that this small publication will inspire folk around the Globe to take another step.
I believe this is the first time that a Large, 6kW, 48vdc to 230vac, Pure sine wave 50HZ, Inverter has been manufactured and open sourced in this way.
However 'Phred', These forums have everything discussed and required to make a truly remarkable OzInverter, but remember no process should be skimped.
The OzInverter... summary... Get the silicone iron wound cores right, get the hand wound toroid correct, get the connections correct, get the cooling system correct, get the Boards in right...... Do it all properly..... and It will do 6kW for some time, that's if the batteries can take it, but it will also do even bigger/huge 15kW loads for a short time.
-
Phred,
Regards, making and selling boards, Nope, not my cup of tea. I am not in this for scrabbling around for monies of the backs of others, I just want a good Inverter for my boys, our community & friends.
And remember these New Boards are specifically designed for the Big Torroid OzInverter.
-
Thanks for the kind words Clockman ( yes folks I paid him to say all that......)....but
"That's excellent oztules, nice layout, so you have a working Control card with that egs002 card with that 8010 chip.
I do love robustness and simplicity."........
Don't forget you need to disable the 393 and jump some pins, or it will self destruct with a torroid... once done I have not been able to blow it up.... before that I blew up 5 boards trying to find out why...
Everyone has probably guessed already..... that this is indicative of me not having a clue what I'm doing....
...........oztules
-
I have watched these posts with great interest, beginning with part 1 thru part 2 and now for a new more robust board.
There is one constant that I see.
SAFETY FIRST
OZ talks about it throughout his entire posts.
ClockmanFRA talks about it throughout his entire posts it is in the book. AND it continues to be mentioned in this new post.
When dealing with voltages of these levels and possible personal harm. IF it is not followed, it will bite you :o.
Bravo to both OZ and ClockmanFRA for this.
Cheers
Bruce S
-
Note on using a solder grommet through a board. It was a failure point in older electronics that used them. I remember having to drill them off many boards then patching traces and adding a wire top to bottom. They would corrode and break in the center...
-
Thanks Mary B, I have beefed up the pads on both sides, so some good soldering both sides for the component legs should give a solid joint.
I use 2off those 1mm dia pins per track joint from one side to the other.
Also beefed up the FET legs where possible, and ensure they do not get to close to things. The PJ's get a little scarily close in places ::).
Below Track Mask for component side, the final mask will loose the centre points and script, but the six Board mounting holes will remain for correct alignment of the other side Mask.
This a JPEG image translation from the CAD file so its not so good. The CAD file print out is very very accurate and in World dims, so its up to the Laser printer now, lets see how good or bad they have become, when it finally gets here.
-
At least use something like GIF or PNG to get the image our of CAD and to your printer if you can; JPEG is not designed to work well on sharp edges and small numbers of colours.
Rgds
Damon
-
That's the CAD's direct conversion to jpeg at 100k, PNG & GIF were even worse.
I will do a inkjet print out from my normal printer using the CAD's own printer software, and then scan it in to the PC at 600 and size down to below your 512k image max size.
Bit of a fiddle, as the Power board, although fits on a sheet of A4, at real World scale size is 5mm to much for the printer. So I have to tape on 10mm of paper strip at the front of the blank A4, tell the printer its a Legal 8inch by 14inch size, then tell the CAD software the drawing is A4 and now 310mm long. That works.
The new Laser printer will be interesting.
-
I'd recommend Post Script (or even pdf) as most laser printers can swallow it natively. Its a vector format so no fuzzy edges at all.
Certainly worth a try - perhaps try a couple of prints side-by-side.
-
Cheers frackers.
Here we are...... remember this is first design layout on the component side, but there may be alterations.
For instance, top left FET left leg to copper area is better than the original PJ board, but it still concerns me. Especially if you consider that the copper is relatively thin and its a Power Board. More copper the better is my opinion, so I may move those FET's and resistors along to the right 5mm. But await to see what 'oztules' recommends.
-
Use a heavier copper board like this 4oz board with much thicker copper http://www.digikey.com/product-search/en/prototyping-products/prototype-boards-unperforated/2359509?k=&pv46=19614&FV=2000017%2Cfff40024%2Cfff800d5%2C11800a3&mnonly=0&newproducts=0&ColumnSort=0&page=1&quantity=0&ptm=0&fid=0&pageSize=25
-
@clockmanFRA & friends:
Don't want to rain on your parade but how "open source" is a design copied from a commercial product?
Furthermore, what about this "open source" project when the original manufacturer quit the production of the control board (the inverter's "brain")? Do you have its microcontroller software?.. is that "open source", too?
What if I do reverse engineering on the Windows 10 code and only change the default font (as you changed the original transformer with a diy one)? May I claim it the "open source" OzWindows?!
As a disclaimer, I've made my own 6kW inverter from scratch (including all the software) with absolutely no commercial "influence". Actually, the schematic is pretty simple but I came to the actual variant after months of testings.
By the way, the PJ doesn't use a state-of-the-art design when it comes to Mosfet drivers. They're rather rudimentary and a more up to date design could further improve the inverter efficiency.
Good luck with your "open source" project but you seem to not really own the whole "source".
For that matter, any electronics project should start with a schematic not a PCB. ;)
But you're just adorable when you talk about "changing 47R to 5R6 and put a reverse diode" without knowing WHY or IF that's really the optimal choice.
-
So glad to hear you helping people lifer.
Yes the circuit is stupidly simple now.
Interestingly, power jack have incorporated some of my changes into their units, because they are better that way.... so must read this too.
But yes your criticism is valid, in that this older stuff is captive to PJ continuing to supply their boards, and it is to this end we have now gone to firstly the 002 board for a quick test, and now to a simpler board using three chips and a few components.
This appears to be not rocket science, and the next step is a pic chip, to replace the 8010 chip.... maybe then you'll be happy, but I need to crawl first... I'm not an EE.
It is fascinating to see how secretive the run of the mill EE is, and how they desperately try to keep everything out of the reach of simple folks like me.
It's not working... we are making impressive units with first the PJ parts, now eg002 parts and soon just the 8010 chip and soon only off the shelf parts.... your smug air of arrogance will have some wind taken out of it then.
When you publish your hard won units circuitry and software for all to use and copy, then I will have some respect for your snide comment, and I will design anyway I so choose.....
I happen to like the suck it and see way, and so I design in the pcb progam... perhaps your unable to do that, and need to spice it first as you don't trust your designs in your head... I don't know and don't care, I will do as I wish.... and what hurts the most..... is I get results, and when that s not quite right, I modify and try again.. thats how I learn.
For now your just another chair critic with nothing to show for it, and unhelpful on a diy site. Your certainly able to look down on us, but thats not the view I see from here.
Remember, I am the village idiot, so I have nothing to loose by trying stuff.
Here for your merriment is my latest test bed for the three chip version.
And yes there are almost no components it is dirt simple...... it will hurt you I suppose if it works as well as the previous 002 boards did
[attach=1]
Will it work as expected... I have no idea, but I will find out... and it will eventually work as expected. You can help if your able.
Should I expect to see documented evidence of your trevails?
..................oztules
-
If i may offer a little input... If you were to add a parts list and schematic for 24 volt.... Then poorer people can start at 24 volts and upgrade as they can afford to by making a new 48 volt board/or replacing parts to re-vamp their system from 24v to 48v. I guess you could even include a parts list and schematic for 12v and 36 volts.... Really make the system fully scale-able..... Grow their system as affordable.....
Just saying..... Frank
-
Hi frwainscott.
The 48v present version is because I had a need.
Generally in DC terms most controllers are 48v, and more importantly a 48v dc battery system is a better cost effective way forward, and at even equalise voltage of say 61v my boys would not be harmed.
Maybe some one here can take the 48v and modify down.
Midnite Solar, (USA) are at this moment bringing out a new range of EGO Inverters, and 3 types of Wind Turbine Grid Tie Inverters. So that will be interesting.
-
Hi MaryB, and thanks for that.
I will definitely have a go with those in your links, along side normal boards and watch what happens.
The New boards will be going on the BigOzInverter so that will really test stuff.
-
@oztules:
I don't want to hijack this thread, but I owe you some answers:
I might have showed up my design if these were all about "open source", as a result of a big brainstorm of many contributors and such. But the actual project is just a clone of a (outdated) commercial project thus it's not smelling as "open" at all.
And yes, the inverter design could be stupidly simple. I'm actually using an arduino-like board (more like the nano form-factor ones), four optocouplers, four performant Mosfet drivers (Microchip TC4422A), 4-6 power Mosfets per switch, four big (20uF each) polypropilene capacitors as a DC-link, a handwriten (old school) PCB and.. that's all.
Almost no other discrete parts (except the gate resistors/diodes, bootstrap and decoupling capacitors and so).
I've read tons of datasheets and application notes thus I was truly understand what's going on and how it actually works.
It's a game of nanoseconds, if you were to chase for the last bit of performance. You can't just switch a gate resistor with a random one without knowing the Mosfets parameters, the microprocessor generated deadtime, the PCB parasitics and so on.
This inverter topology is a hard switched one thus you have to minimize the effect of body diode conducting & reverse recovery. Too much of a deadtime could lead to higher spikes acros Mosfets drain-source due to body diode reverse recovery current. Too small of a deadtime could lead to shot-through. That's why I was saying that playing with this times is actually a bit of rocket science.
Of course, you may go the "destroy and learn" path but I'm a too big Mosfet-lover to watch them smoke.
I have no problem with no-EE people but it's a common sense to not write books about a subject you don't understand.
Btw, I'm not using spice & stuff as I'm only using linux and never bothered to play with wine just to simulate an circuit. I have some schematics drawing (and PCB design) software installed but I've never seriously used them.
In fact, I'm hand-drawing PCBs for almost 40(!) years now thus I come to "feel" the traces.
Especially for this kind of sensistive boards (high frequency, high currents) it's much better to manually design the PCB to avoid many culprits (parasitics, ground loops and such).
And yeah, I'm using single-side PCBs and through-hole components! I'm not against technology but I just have enough space for this kind of projects (and I can't manufacture vias anyway).
-
Hi Lifer,
@clockmanFRA & friends:
Don't want to rain on your parade but how "open source" is a design copied from a commercial product?
Furthermore, what about this "open source" project when the original manufacturer quit the production of the control board (the inverter's "brain")? Do you have its microcontroller software?.. is that "open source", too?
Firstly 'oztules' has been discussing these OzInverters for nearly a year now, about 80,000 words now, he experiments in the real World and makes stuff work. In my book and in this day and age any one that really does, shows it and openly talks about it should be heaped with praise.
I have followed his adventures and built my own OzInverter with 'oztules' guidance. He say's things as they are, and in a understandable and logical manner.
Stumbling block for me was that I need real power for our community, we were badly let down by SMA, and the latest Chinese Power Jack boards are now difficult to obtain for my next BigOzInverter.
I new 'oztules' was playing with that new chip and the EGS002 board, because even I mentioned it.
When he said he got it work okay with toroid's by getting his Scalpel out, then I got on with the NEW Power Board. Then 'oztules' said he was doing a New Control board with just that 8010 chip....................
'Open Source' I have discussed that philosophy earlier in this post.
'Open Source' with the New Power Board I am doing.....
I purchased from China an old stock simple PJ board. Its a very simple layout each FET gate has its 2off resistors like all FET's have.
However, with the 8010 chip that the New OzControl board has, then the Power Board I am putting together
has to have some simple modifications.
Is my Power Board a copy of the PJ board, No, in my opinion it is not, its superior.
I have sorted out a few things along the way. I also doubt that my New Power Board would even run in one of PJ's Inverter units.
So 'Open Source' to me...... means...... start from the ground up, keep it simple, re configure to do the job, see what's out there and incorporate what is good, use the latest readily obtainable bits and pieces, and importantly be cost effective. Then post, tell, publish what's being going on.
I have a need for an Inverter, the Big SMA let me down, I made my own, now I am making my own control electronics. With 'oztules' help off course.
I am not specialised in the EE field, so I can come up with wacky ideas without feeling a fool.
However I do understand Safety, Care, and due diligence, but how far does one go on that?
'Bruce S' has already mentioned this. And I certainly say it when I 'Open Source'.
Ps, I name stuff the OzInverter, and the OzControl Board, why, because he does it and tells the World, brilliant.
Credit due, where credit is due.! :) :)
-
@oztules:
I have no problem with no-EE people but it's a common sense to not write books about a subject you don't understand.
Oh dear......
What I have so far done and made, is now published. .............................
SMA again?
-
I'm also living offgrid (and that's what actually forced me to study the power electronics in detail).
Once again, I have no problem if someone hacks a commercial product to suit his needs but just don't call it "open source". It's not what open source is all about.
Actually, you don't even have to look at PJ boards anymore. Any semiconductors manufacturer has dozen of app notes which describe in detail this kind of inverter topology.
And like I've just said, you don't need a "black box" (aka control board) either. A simple arduino-like board could do everything a "black box" does (and much more).
An open source inverter project should start with a complete schematics. Where is it?
You just published a book(?) about hacking a commercial product (actually, was all about that transformer replacement). Why not posting that pdf on this forum and the similar ones (anotherpower, energymatters, backshed etc (which were actually the real collective author)? Why do you need to keep it secret (printed) and only making it available (even free of charge) by post?
That's not "open source" for sure. With all due respects, I guess you just like to see your name on covers by any means.
Speaking of that transformer hack, it actually started long time ago and it's author was an oz fellow too (andrew).
But it wasn't rocket science either (anyone with medium electronics knowledges would have come to the same conclusion). That's why the real author didn't even bother to write his name on starts for that.
I'm not against open source (in fact, I only use open-source software and even contribute to some) but this thread has nothing to do with it
Speaking of SMA (or any other top manufacturer in this field). Why haven't you cloned one of their products? Just because PJ it's chinesse/ship made gave us the right to hack it and claim it "open source"??
-
@clockmanFRA & friends:
As a disclaimer, I've made my own 6kW inverter from scratch (including all the software) with absolutely no commercial "influence". Actually, the schematic is pretty simple but I came to the actual variant after months of testings.
So where is it? Is it open source? Where on github or bitbucket do I find both the software and hardware details?
If what you have done is so much better, why are you keeping it a secret?
-
Lifer, if you think your design is solid, then you should put it out there and help people, and do a better job than me... and thats easily done.
As it is, all your research, and scrivenings will die with you... big woop.
Not everyone lives in first world countries, and would appreciate open honest work that may help their every day lives.
The open source thing is for others to ponder, I just put my work out there, with all the pimples and bruises, warts and all, tell what I know, how i have done it, and no-one has helped me to be any wiser.... thats the calibre of the other knowledgeable folks out there... a big fat zero.... and your just another one until you prove me otherwise. I welcome you to pick it to bits, advise where it could be made better.... put your better deign out there, and help the other folk to improve their lot.... I don't expect it to happen, as selfishness seems to be the order of the day with most people.
This site was made possible by the selfless experiments of the Dans. They did plenty wrong for plenty of time. Great people like Flux and Zubbly contributed to their knowledge, as well as the best teacher of all experience... they are now world wide acknowledged as leaders in this field... even if I don't agree with the Hugh piggot/ Danb mills as the best long term... I think that prize goes to Zubbly.
I built a few of theirs, and they did what they claimed ( better in fact ), but my marine environment made the neo's die fast, and the furling was difficult to control in some wind regimes. The back MMF protection of the zubbly made it near bullet proof, at the expense of output perhaps, but the best in my..... thats my book.
These little 002 units seem indestructible so far, and it is not that I haven't tried.
I do understand the gate capacitance as it pertains to oxide thickness, and surface are, and low Rds units have bigger surface area and so bigger capacitance to over come if you need to switch quickly. ... maybe even thinner oxide in the low voltage variants....who knows.. not me.
The 002's are running at 300ns dead time, which seems to cause no conduction issues or at the other end, heat and excessive waveform distortion. Lower resistance was chosen to increase the current to the gates, as the 6x 4110 has about 6n of capacitance to drive into ... lower impedance means faster turn on times, the diode to help decrease turn off time.
Interestingly it does not matter really what the theory might actually be, practice says 47r runs hot 5r runs cool.. why it does is of no importance, and you don't need to test to destruction to find out either. kiss works.
So carefully thought out does not necessarily mean better or worse, but if it works to expectations, then I don't care if I followed your example, or went my own idiotic way, the result is a unit equal or better than other units I have to fix for other people with name brand goodies.... They can't drive anything I cant, and I can drive a lot of things they cant...and for longer periods. ie I can start 10hp three phase motors for phase generators to drive 50 ton presses.. theirs cant stand the start up because of the programming probably, but the reason is mute... they can't do it... I can
I don't need to know any more than that.
I have never pretended or expressed that what I do is open source, thats for other folks to ascribe words to, read any or all of my stuff.. I don't recall ever mentioning it... but what I do is open for all or anyone to use or modify at their leisure.
I am happy that clockman and many others now have had great success using what I have done to further their pursuit of useful power.... no one has used yours from the sounds of it but you.... and your happy with that...... your loss as a human being not mine.
I had intended to use the 8010 chip with a triple isolated pwm supply, and drive optos, but have not moved forward with that one as I was pleasantly surprised that the 2110 can drive the 6 fets very very well, which I had not expected... provided I dropped the gate resistance.
You can laugh at this one too.. I know the PWM part works fine as I used that for a three phase BLDC driver for solar pumping I did a neighbour...... never went on with it though.
[attach=1]
I will end with this... if your not prepared to help then go away from this thread, and grizzle to someone who cares.... or you can pitch in with sensible critique... there is no middle ground.
...................oztules
Edit Saw your last post... Yes, Andrew did do a piece on energy matters. Thankfully I did not see that or I would have ended up with a less useful device. He calculated the inductances to use that would mitigate the idle current issue.
In fact he was wrong. The inductance was way too high, as his friend found out after reading my saga with it. He ended up using mine instead.
Had I seen Andrews, I would have made the same mistake/s I guess. His filter caps were wrongly ascribed as current hogs to. The originals were just fine.
And who in their right mind would want to replicate SMA.
-
It would be good to take some of the heat out of this thread.
In particular I'd say that open source means lots of different things to lots of different people. I tend to put my stuff in public repos (or at least downloadable tarballs or similar) under permissive licences such as Apache/BSD. Others will say that anything other than GPL is selling one's soul to the devil, for example.
One annoyance that's I've experienced and may be a factor here is that I've put up some work (photography) to allow others to *use* it freely, but I've caught people watermarking it as their own creation and trying to resell it, which just reeks of dishonesty to me.
So, deep breaths all round please.
Rgds
Damon
-
frwainscott
"If i may offer a little input... If you were to add a parts list and schematic for 24 volt.... Then poorer people can start at 24 volts and upgrade as they can afford to by making a new 48 volt board/or replacing parts to re-vamp their system from 24v to 48v. I guess you could even include a parts list and schematic for 12v and 36 volts.... Really make the system fully scale-able..... Grow their system as affordable"
The current 002 board variant I am testing until my other chips arrive is already scalable, and multi voltage.
The only difference with a 24 or 48v is the transformer primary, and if you wind that two in hand for 24v, then 48v can use the same transformer by placing the pimary in series instead of parallel..... ie you will be 16:1 for 24v and 8:1 for 48v for a 240v system... nothing else changes unless you need better fets, which I doubt unless your really going for broke... and at 24v I don't think that wise... the currents become unmanageable realistically over about 3kw or more.
12v will require a small boost converter for a few dollars to get the voltage up high enough to reliably provide for the gate voltages in my mind, other than that it is fets and transformer again.
I may acquiesce to the circuit thing when I decide what design will be the best/final thing... until then I will do as I have, and only provide what I have from current experiments and that does not include complete circuits... some one else can do that if they wish.
Damon,
Anyone is free to use, steal, pilfer anything I have written here, as it is owned by the site after all as I understand it.
Courtesy would dictate that it may include a reference to me, but I won't lose any sleep if they don't. Most folks have given some form of recognition as far as I know, be it chainsaw blades, alternator repairs, solar panels or a host of other things.
I thank Clockman for being up front too. Is it open source??.... don't know and don't care, so long as someone benefits somewhere I'm happy.... .. obviously, not everyone is like that.... which is just messed up.
..............oztules
-
My two cents ,
if the product comes from China it's most likely copied from some other manufacture anyways.
-
Yes Neil, there are dozens of chinese inverters there on alibaba using the same generic boards last I looked ( few years ago, before PJ agreed to sell me theirs as replacements for my unit I bought... which I promptly blew up.... had a W7 also ( same thing, better build, but EI transformers... blew that up too ( spider on circuit board))
Then I asked PJ to put them on Ebay, and they did, and sold lots of them.
I don't think they have the software, as they still havent solved the transfer problem with charging and turn off with torroid transformers. That software works in the W7 just fine, but the lack of leakage in the torroids kills the fets dead, depending on the part of the cycle it was on before disconnection.. if near zero cross over.. then death.
..........oztules
-
if the product comes from China it's most likely copied from some other manufacture anyways.
There's no honest people living in China, huh? After all, to steal from a thief it's a honest job, isn't it?! Okay, let's just pretend you only "assumed" the PJ it's a counterfeit product..
Nevermind, I wont continue this discussion neither. If someone is really interested in spreading inverter theory (and step by step diy instructables) to the masses, we could open a different thread, to post full schematics and source code so everyone could customize its inverter.
I also thought of these inverters as being some magic "black boxes" at the beginning (due to their prohibitive prices and lack of informations).
Now I can assure everyone that it's actually pretty easy to build one (even if you don't really understand in details how it works).
But, as @DamonHD has corectly pointed out, any honest (hardware) open source project could be easily replicated by any commercially minded person without a notice.
We could spent months of tweaking and testing to get the best results and he just choose a catchy name ("PowerJack"?!) and put it on production (and eBay)..
Now you tell me: are we helping the undeveloped countries or the nasty ones?!
-
As a little diversion, here is a modular MCU-driven inverter project:
http://sustburbia.blogspot.co.uk/2015/10/a-micro-solar-inverter-based-on-arduino.html
Rgds
Damon
-
it is what it is Lifer
-
Lifer, I see you have a bee in your bonnet about something......
If you have not seen my little publication how do you know what's in it.
Its about me making a OzInverter, and throughout the publication I always give credit where credit is due. And these forums are prominently listed.
Yesterday, I gave a copy to an electrician in Morocco who's home village has no Electricity, and he will have a chat with his nearest college elect dept to see if they can implement the OzInverter.
Last night I gave a copy to a French Doctor who comes from one of the French African colonies, and again, "Thank you I will give this to our local college at home and they will take it forward."
These 2 were truly excited at self empowerment and self control and not be beholding to costly incompetents of others.
That's what I mean by getting the Information out there. There are very intelligent practical people around the Globe that can truly benefit, so with all the information in one place, a book, with lots of pictures and written by someone that has trod the path, Me, I can see things from there point of view...............
I will not argue with you anymore 'lifer', as you have preconceived ideas, that ........
-
@clockmanFRA:
You don't have to give it to an electrician or a doctor - just give it to the whole world. Why don't you just share a link to this board??
What's your book about anyway? About taking a commercial (obsolete) product and changing its transformer? Man, I should write a book about changing a light bulb, too.
If you really are open (source) minded, just put your pdf here. Otherway, remove that "open source" label from your sentences.
Now if I'll start a new thread with a REAL pure sine wave inverter project (complete, scalable, easy to customize and diy friendly), would you write a book about it, too??
I really-really want to help people to build such an inverter from scratch (or a battery charger for that matter) and I'm seriously thinking how to gain control of all those informations.
If I post them on a board like this, I'm afraid I can't stop a scaveng.. umm, "writer" from "sharing" those infos to an electrician from Morroco or a doctor from Cote d'Ivoire..
Once again, just think about it: your book is using @oztules "font".
PS: How do you define the "OzInverter" anyway?
-
Cool it please everyone.
-
DISCLAIMER!
I am not in a good mood, I haven't been since 10 minutes AFTER bombs went off in a beautiful town in Brussels! AND I started checking on friends I knew to be traveling.
Then I get a MOD's notice that THIS STUFF is going on >:(.
GM MODE!
LIFER! STOP !
IF you had posted a protest about the "Open source" 6 months ago it might have been valid.
AFTER a full blown build and going on a 2nd, YOU do not have a valid complaint.
One more and you get to be in read only mode .
PERIOD!
End GM MODE
Don't like it? have an owner get in touch with me.
-
followed the postings of clockman oztules operahouse and all the others that contributed
to putting this together. Sharing their knowledge and experience is of great value to me.
Thanks Ontfarmer
-
TY Damon, some pretty condescending attitudes were coming through.
It would be good to take some of the heat out of this thread.
In particular I'd say that open source means lots of different things to lots of different people. I tend to put my stuff in public repos (or at least downloadable tarballs or similar) under permissive licences such as Apache/BSD. Others will say that anything other than GPL is selling one's soul to the devil, for example.
One annoyance that's I've experienced and may be a factor here is that I've put up some work (photography) to allow others to *use* it freely, but I've caught people watermarking it as their own creation and trying to resell it, which just reeks of dishonesty to me.
So, deep breaths all round please.
Rgds
Damon
-
I was reading this forum on a daily basis since I signed up (more than two years ago). I want to thank @Flux and @DamonHD for their kind support.
You may delete my account.
-
That was just gratuitous, "I'm all that", unnecessary troublemaking.
Why? I don't care.
Fragile ego's have trouble with apologies, I guess.
Onward.....
-
Oztules is the best thing to happen to this forum since Zubbly.
Dave
-
The thicker copper traces will change the inductance of that trace, might need slight tweaks...
Hi MaryB, and thanks for that.
I will definitely have a go with those in your links, along side normal boards and watch what happens.
The New boards will be going on the BigOzInverter so that will really test stuff.
-
No Oztules is not the best thing... for one thing, the test board as shown is the one i am using, but forgot to put in 2 changes to it
I still advise NOT to use this pattern, as it is very sloppy in track clearances and placement is difficult compared to the real estate it uses.. but some won't be told, so here are blow ups of the board as it was.. but with two changes.
When I first tested it, I blew up the voltage regulators... as mentioned somewhere back there.. it was because the zener went to the wrong track by 2mm. Changed that, and the voltage feedback didn't work because of no ground link.... fixed both in a few minutes, and it has since been running the house till now... more than a week I think... works flawlessly for a botch board.
It was not until I took it apart that I realised the 2 changes I forgot about in the excitement.... as I had not touched it for all that time.
So here are the blow ups of the changed board.... no other improvements... just the two changes... I would prefer you not to use this until we get a nicer more robust version ( physically), but some folks can't wait apparently.
[attach=1]
[attach=2]
[attach=3]
[attach=4]
That will help those who can't wait for the experiments to conclude.
...............oztules
-
I'm just amazed by all the boards that are sold on ebay without a bit of documentation. I haven't noticed any documentation listed here to that 8010 board, forgive me if I missed it. Here is an overview of that board that may be helpful. http://www.egmicro.com/download/EGS002_manual_en.pdf
-
howdey all,
I have been reading this forum for a fair while and I had to register so I could make this post.
lifer where do you get off buddy. clockmanFRA and oztules have put so much into this community by helping people
with their questions and most of all sharing their knowledge. and you just mosey on in and start trying to belittle their
hard work, good on ya you muppet. fair enough clockman may have made a printed copy of a book and that may not be the best way
of publishing a book in this day and age, but guess what HE DID IT WITH GOOD INTENTIONS and is giving people some help.
the day you write a book you can distribute it however you wish. but until then I suggest you worry less about the term
"open source" and worry more about your wide open mouth.
-
Further testing on board no. 2
The weakness in this version of experimental board is the rudimentary power supply. The 7805 is prone to failure in further bench testing.... and that takes the chip out.
It didn't happen in the first board as I must have been lucky at turn on. In the pj board, the caps must have slowed the inrush currents enough to protect it, but a second one on the bench into a small transformer with no filtering failed a few times on start up ( instantly) and took a few boards with it.... so I have run out of boards for the time being, and won't risk the one on the big transformer, so bench testing will stop until new victims arrive, and a new board with better power supplies and other changes are done.... obviously tried to make it too simple.
I'll keep the big one running until they arrive, just to continue the power testing.. but it looks strong once up and running so there is much hope.
in short don't do that board unless you better the power supply.
The next board will have plug in chips so I can be more gruesome in the testing of it.
Lifer was right, I don't know what I'm doing..... but I will get there for sure.
................oztules
-
---GM MODE---
dr_zogg;
This is your first post ,,, so I will be nice for this time.
Let's NOT continue to go after a fellow forum member, even if he in his most recent post asked us to remove him.
Further posts attacking a forum member will result in a read only status.
---END GM MODE---
-
6kW-15kW new Power Board PCB....... the Mask for the Drilling Holes etc.
-
General layout.....Component Side
(I will do the other masks later, must go do some Taxi the boys). ::)
-
I fail to understand the 5V issue other than it can be tricky dropping from 45V. Just what is a 35C? Given the world of counterfeit chips, 18V is probably a bit much to put on a 5V regulator. I've thought about ordering some 12V 100V>12V open frame wall wart type switchers available for a couple bucks for testing but haven't had a project that needed any quantity. Most 100V wall warts work well at 50V with reduced load. All my shed control electronics works off wall warts.
-
The PCB Component side Track Mask....................
-
The PCB Under Side Track Mask..........
-
The Component outlay & values Silk Screen/stencil Mask...........
-
Couldn't wait.... so rebuilt one board out of three dud 002's... and were trying again.
Found the hairline crack that caused all the snafu's.... not happy about it, but there it was... if found sooner would have saved a lot of swearing.....
In the quest for power supply stability on switch on, decided to make it tip35c for both 12v and 5v with 120R and 100uf capacitior to dull the start up spikes.... seems to have quelled all resistance from that quarter.
If the damn thing didn't work so darn well, I would make the triple pwm supply and get away from the 2110..... but it works so well, I can't justify it on performance terms....
Anyway, second board is working as well as the first now, and waiting for more boards to come... see how many I can destroy next time.
The hairline crack was a bitch, and the board I know is experimental, but I wish I had spent more than a few minutes drawing it up... those tracks are too thin for my board making skills, and the pads too small..... needs a resign for experimental part 2...
Looks good so far Clockman, the diodes should be shottky ( frackers recommends 1N5819 ).
...............oztules
-
7805's do not like high voltages... go over 18 at your own risk... make sure you have filtering caps around the 7805 also. I don't remember the exact values but both input and output need them.
-
Thanks Mary... I've dumped the 7805 for the moment, and running tip35c...
As an aside, are you on the air, and if so when ( ish) ( I was asked to ask you from VK7.... )... or pm me if you want to do that instead
.................oztules
-
I am on the air but fighting severe power line interference. Like S9++ on HF and when I point at it on 2 meters it is S9+60. In a bad spot too, where my moon rise is so it kills doing EME.
-
Oz,
Referring to reply #39, You are absolutely right. Thanks.... I was focused totally on the egs002 circuitry and was not even putting any focus on the fets and circuitry beyond the egs002.
I hope all will excuse my ignorance as i am "LEARNING TO FLY".
-
This is the wrong thread I suspect, but for completedness here is the latest offerings...
I have moved on from the 002, not because it is no good, as it is a ripper really, but I wanted a non imported board, and settled for the 8010 chip.
So now we have three plug in chips only to replace if the worst should happen, ( it does not look likely at all I might say)... board looks like this
[attach=1]
It looks like this
[attach=3]
It works flawlessly, at up to 12kw surges, and 4kw cont... remarkable, wave is good, voltage stability from zero to a dead 4kw instantly has only 1 volt max variation.... yes it is impressive.
The 8010 chip costs about $1.50 AUD..... not worth burning your own... I can't buy a blank for that.
...............oztules
-
The 8010 chip costs about $1.50 AUD..... not worth burning your own... I can't buy a blank for that.
I get the impression that the 8010 chip comes in several flavours - how do you ensure you get the right one? Are you lifting them from an EGS002 board?
I'm in the middle of the walnut harvest here (probably another 3 weeks) no not much progress on getting a logic analyser onto the 8010 on a 002 board to see what makes it fly!!
-
talk to the seller i suspect..... yes... getting one off the 002 works too.
..........oztules
-
Hi
I'm italian hobbyst and i'm interested to the power sinusoidal inverter.
In the past i made some small modified sinewave inverter at about 500W with a stamndard transformer.
Now I want to build something bigger and whit a sonusoidal output (so no noise and no problem with induction motors).
I need a 6KW power.
I have already a toroidal transformer 220V at the primary and 32V at the secondary with 5000VA power.
I have some EGS02 module too.
I need the complete schematic of the control board, with the bill of materials and tha same for the power board.
I am aware that I am asking a lot but I want to build something much better and i want to risk.
Can you help me please?
If you want you can send me the schematic to my email: giordanto1971_AT_libero_DOT_it.
Thanks in advance.
Best regard
(forgive my bad english).
Antonino Giordano
Antonio, It's not a good idea to leave your full email address on here.
AND it would be better to ask via PM messages.
-
Everything you need is in this thread.... http://www.anotherpower.com/board/index.php/topic,1116.0.html history, failures etc... inc pdf of boards and someone has done a circuit as well I think..
This is the relavent post perhaps
http://www.anotherpower.com/board/index.php/topic,1116.msg11614.html#msg11614
Here is a circuit for the power board...
[attach=1]
The resistors need to be changed to 5r6 and reversing diodes placed across them.
Clockman has provided the screen for it.
...........oztules
-
http://www.aliexpress.com/item/10pcs-lot-New-EG8010-LQFP-32-Sine-wave-inverter-chip/32605461001.html (http://www.aliexpress.com/item/10pcs-lot-New-EG8010-LQFP-32-Sine-wave-inverter-chip/32605461001.html)
The 8010 chip, that has the correct numbers...... 1536 at the end.
I have ordered and they are on the way. ...... we will see.....
6kW-16kW Power Boards..... The only sticking issue at present is the heat sinks, one is 271mm x 80mm x 26mm and the other two are 131mm x 80 x 26. These are coming from China, what we want, and good low cost, but, been nearly 7 weeks now.
These are important as I might need to redesign the mounting positions as the long heat sink makes the big PCB Power Board ridged, important with those 9off 16mm/2 flex cables secured to the heat sinks. Vertical upright mounting can not be beaten for simple heat dissipation and simple fan cooling.
On the Power Boards themselves I have allowed the Solder mask around the FET legs a lot of space for extra solder etc, (straps if needed, but need to test at 15kW, approx 300amps).
The real size board masks that will go with the, "Making a Low cost, simple and Robust 6kW, 50HZ Pure Sine Wave, 48vdc to 230vac, INVERTER." book as a pull out, will have the option for solder masks as folk might not have the option for double copper thickness on their boards, so will need to beef up the copper.
I will also do the book with the option of also purchasing the bare PCB boards, that's the '6kW-15kW Power Board' and the 'OzControl Board'.
Just finishing the 'OzControl Board'.
'OzControl Board' ....... again Credit where credit is due.
And this Old Fool will need to ask 'oztules' a couple of questions ?.
-
For those in australia using the inspire inverters... we get massive heat sinks... so I built one today like this
http://www.anotherpower.com/board/index.php/topic,1116.msg11763.html#msg11763
..........oztules
-
HI
thanks for the quick response, it is a real pleasure for me to ask for explanations and receive them.
I will start to work on the inverter as soon as possible.
if I have problems I will contact you.
I appreciate your cooperation and your explanations.
regards
Antonino Giordano
-
Oztules....
Question please.
How does the 8010 chip handle the back charging, ie, my Grid tie Inverters sending there power through the Ozinverter to the batteries, that's if the Grid tie power is not being used.
At present the PJ arrangement allows a slight rise in the 230vac to 242vac then the GTI's shut down, or the safety relay activates if the batteries DC charge controllers activate to Float mode.
Just not sure how the 8010 will regulate itself.
-
Good source for heatsink extrusions in the USA http://www.heatsinkusa.com/10-080/
-
Clockman.
It is a function of the synchronous switching of the H bridge, that power will go both ways... so by definition backfeeding will be the same as the PJ in every way.... no control over the battery voltage unless you control it with a separate controller to turn the gti of when batts are full.
Nothing will change for you.
The charging by the PJ in the official mode is controllable, but the software knows whats going on then, and acts accordingly, but not with GTI.
..........oztules
-
Thanks.
Reason for asking ........ Now that i am actually doing your OzControl Board i notice that the Chip wants to see the AC, well the rectified clean signal, on Pin 13, 14, and 15.
I therefore presume this is for HZ stability, and will not be effected by voltage rise. Thanks.
I see you dumped some of the earlier Cap cans 4u7, ? replaced with polarised Tant beads?
The 10k and 2k trimmers, what result are we looking for here?.
Mary B,........... sadly the fins are in the wrong direction. This is what i am after so folk can have a steady supply........
-
Huh? Lost me.. they have many other heatsinks and you order by the inch! http://www.heatsinkusa.com/
-
No Clockman, they are these
http://www.ebay.com.au/itm/221465112357?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT
monolithic ceramic chip... small, and non-polarised.
pin 15 is temp shut down/sense... depends how you want to use it.
pin 14 current sense input 2k voltage divider from CT
pin 13 voltage sense input 10k voltage divider from 240v ac output.
Voltage sense is to keep voltage at 240v or whatever you set the pot for.
current sense is for shut down at whatever level you set.
frequency is not sensed, it is set by jumpers etc.
...........oztules
-
Yet again, thank you Oztules.
I see you are using a temp probe from one of those button cell thermometers for temp shut down,? is this effective? is the temp as per spec of the Chip 45.c ? seems a bit low to me?
One more very foolish question...... that most folk on this forum will fall about in hysterical laughter.
What are you actually doing as a ON..OFF...switch/ latching button with this 8010 board/Inverter.
PS, I do try to ask these questions on your EGS002 discussions, but censorship of some sites by the authorities here, is not getting any better.
-
Okay, On /Off is operated by allowing the voltage to rise to above 4.3v, ie, 5v by closing the circuit and Pin 15 on the 8010 HT1536 chip sees this as an over temp and shuts down gently.?
The New Power Board and OzControl Board coming together.
-
Looking good.... now with just off the shelf components, we can finally make a real inverter..... that is easy to fix as well... utopia indeed for remote living I don't think any other unit on the market can say that.... usually you have to replace propriety boards... if you can get them....
" Pin 15 on the 8010 HT1536 chip sees this as an over temp and shuts down gently.?"
No it shuts down instantly without causing the torroid to blow things to bits, and always soft starts from there.... we can use the inhibit pin on the 8010 to get the same result without the flashing lights.... but works identically.... the temp sensor can still be used without any change as well.... ie when switch is open, the temp ntc takes over the circuit.... use as a over temp shut down, and use our own fan control for the fans.
...............oztules
-
Just wanted to say thank you to you guys for doing this; - whilst my powerjack has works flawlessly, it'll be good to have an inverter that's diy repairable without spending £100's for spares. :)
I kind of feel I ought to have contributed, but even if I'd found the time, I doubt I could have kept up.
-
mab;
Once I saw how quick and far OZ was going!
I decided it was best to grab a good cold Porter, sit back and learn!
Cheers;
Bruce S
-
Clockman, I didn't see this for some reason
http://www.anotherpower.com/board/index.php?topic=1116.msg11784#msg11784
Note: the 7812 and the 7805 are actually tip35c transistors.
Here is a screen shot of the pdf I posted as well
[attach=1]
It does not affect the board obviously, but if you build it using those (7812, and 05 regulator )parts....... you will have problems... ie no power and dead regulators.
........oztules
-
Thanks for clarification Oztules.
Those 13v and 5.6v zenners, 1 watt okay?
Only had a few boards made, and I will probably do a better Silk screen, like the Power Board "Painting by numbers" style.
The OzControl Board .........
-
The New Boards.
The New 6-15kW Power Board has many improvements and changes.
The OzControl board in the picture is single sided, but I now have a double sided on its way, which is shorter in length. see screen print.........
'Why re-invent the wheel'..... I like the Boards overall and complete unit, as its a strong structure when all assembled, and easy to maintain/get at.
I spent some time ensuring that all the components and parts are readily available and easy to source.
When Mounted vertically the unit is easily kept cool, I use the OzCooling circuit, is just right.
I supply all the real size Etching masks with the book, so the PCB's can be made at home. I will keep the single sided OzControl board just for the Book, as it only needs a few 0r/ jumpers/ links.
According to the French Post Office my New PCB's have been in Australia for the week, so hopefully ‘Oztules' should get them soon, then he can really give them some ……………….. Thanks.
-
The Mk6 OzControl board.....
-
Some uptodate comments and modifications.
The No10 OzControl Board. This is the Pin 6 on/off board, but the on/off cable leads should be very short, about 100mm. However, the Temp sense on/off seems rock solid.
Remember we are using toroid's so the soft start is very important.
A small AC transformer on the OzControl board is now used, as isolation is required as some homes may have a Earth/Neutral connection.
At present work is under way by forum members on the 'Backshed Forum' regards the Short circuit on the AC output side, with a suitable reset system. I am watching with interest and will test and modify my boards when appropriate.
At present I fully recommend NOT short circuiting your AC.
-
Utube Video of the 'OzInverter Guts'.
https://youtu.be/C3If9sbpdcA (https://youtu.be/C3If9sbpdcA)
The most recent Boards.
-
The Double sided, Double copper thickness, Power board.
-
Power Board Componet layout.
All components are on this side of the board.
-
This article has got to be like following the very first beginnings of the Zubbly ideas.
Wonderful, but I am now too old to get into it.
I hope younger people will try and follow it.
Dave.
-
Folks,
I haven't been here for quite a while, as now retired; and, instead of a solid 100 M internet connection to my work desk, I'm now really struggling with an unreliable, very slow (14.4K??) home dial-up connection.
Am not here a lot due to extreme difficulty signing-on ( 1/2-hr. or-so to just get in, then exceedingly long waits for access to this and other sites), and then receiving downloads; so, please don't expect immediate response back to you; though, I will be here now more often than recently. Most frustrating!!!
So, please, if you have replies, no large attachments or photos! At least not for me.
I've been in professional commercial and military electrical / electronics manufacturing / engineering/ product test engineering & by hobby here in the U. S. A. for some 6-decades-plus now, so have plenty of experience. Because of my early hobby interest, I also know which end of a screwdriver to use, not just booklearning! And, I've always been wide open to other's information.
Recently purchased and received Oztules 6--15 Kw inverter book, supplement, and circuit boards; the circuit boards are wonderfully commercially and professionally made!
One question for Oztules and anyone else on here: I'm entertaining using a half-bridge to drive the inveter output transformer in a "push-pull" arrangement into a three-lead split winding primary. It seems to me that this would slightly increase efficiency, and cut the quantity of power-drive components by almost 1/2; thus, making the project even more cost-effective.
Has anyone tried this yet? Elaborate about problems encountered, if any?
I look forward to any and all comments!
Regards, JLB
-
The most glaring reason the Hbridge is better fiscally...
1. The transformer would need twice the VA rating in push pull..... ie you need to wind twice the primary... but at the same thickness .. so twice the copper
2. Because of 1....... the cost goes up by well over a hundred dollars... and you save 12 dollars on the fets, but you need an extra identical primary, and because it won't fit.... you will need a much bigger core... more cost..... and you gain nothing for all of that. ( yes duty cycle is half, but if you halve the copper to make the new primary the same weight as the H bridge, the I^2 R losses will kill you any way.... twice the heat even though 1/2 the duty cycle on each cycle.
Thats the simple best reason I can think of.
Im not smart enough to get into the nitty gritty of how and why the Hbridge is a better electrically......... but the money in the industry says full h bridge for real power.... no one uses push pull for these things. ... although push pull is extensively used for HF inverters level shifting stage, but that seems to be the end of it.
Square wave inverters used push pull, but that was a long time ago, and electronics was difficult, and the simplicity was worth the extra transformer size.
And I think it has a lot to do with natural damping due to the shorting of the windings each time, rather than letting them float each half cycle... I guess that stops the ringing fast.
...........oztules
-
Oztules,
Thanks!
I guess maybe I should sign my salutation as "old-fart" from now on ;D, hi, hi.
Regards, "old-fart", JLB
p.s., Anyone else have any further comment or info., or have already tried?
-
I bought a couple ebay boards and I will be doing push pull using a small UPS transformer. Of course I am only planning it to be 100W. Cheap and simple using the UPS FETs.
-
Good evening everyone.
I recently made the inverter board of 15kw, but I'm having a problem, when I connect the board, I realize that the gate outputs "HO" have increased voltage, which ends up burning the mosfets. I noticed that there is a voltage buildup on the two transformer outputs, returning voltage on the "VS and VB" terminals of the ir2113 drivers, which increases the voltage of the "HO" outputs of the ir2113 inverters. What can this be? Can someone help me?
-
Hi dr. Henrique,
Please see my reply to your EGS002 topic that you have posted.
Any good pics of your boards please?
I have not used IR2113.
The two FETS drivers we use are still the IR2110 these have a power supply of 13v by that zenner on the TIP35. This is a Oztules arangement and overcomes some Power supply issues. I have heard that some of the IR211- drivers come with and without internal blocking diodes or such?
On my particular Ozcontrol boards I have now upgraded the zenner to allow 15v to the IR2110. Several reasons mostly due to the quality of some IR2110 that are on the market. But Oztules has left his at 13v as he see's no issues with his particular component supply.
On other forums, a few builders of the OzInverter have many varied issues and I have watched and listened to there build story's. In one particular case the builder used a modern solder and was not good at soldering and Oztules himself had it posted back to him, re-soldered replaced dodgy components and sent it back as a working machine. Other builders have components around the wrong way, others the wrong values, other have tried to improve our boards and caused other issues.
I thought I had given enough information, supplied at cost, in my booklet and the supplement and my robust painting by numbers 3 PCB's, that most folk would follow !!!!!!!.
Comments.
The OzInverter works, if the instructions are specifically followed and components used that we recommended.
It is important to have good build quality.
It is important to follow the Torroid build and test as recommended.
It is important to use good quality components.
The DC power to the OzInverter should be sufficient to operate and not instantly sag, this also relates to the cabling.
With a large toroid capable of continues 6kw, we need to start the Ozinverter gently and build up, in a manner of speaking, we do not go with the toroid with the accelerator flat out.
The On/off switching on the Pin 6 of the EG8010 sine wave inverter chip is critical and the cabling should be shielded or less than 100mm long, or decoupling with a 100n & 10uf caps bypassing the 1k resistor that goes to ground negative, (basically the off is a real ground, on is +5v but any bounce and flutters and the start up flutters and this is not good on a large toroid and the FETS suffer).
So make sure that your On /off switch is good quality, I use push button latching types. However the overtemp, pin15 is very stable as a start up, ie if it gets more that +4.7v the inverter shuts down and restarts safely when the +5v is removed.
I am, rewriting the OzInverter book, but its an issue for me to get down all salient facts and necessaries without being boring, but then not putting enough information.?
-
When you're ready, send me the link, because I'm going to buy it.
-
I made some experiments using the lay out of the frequency's board, applying to it using a switched source voltage, and I realized it that the cooper plating that are above it, on the other side of the board, appeared one voltage equal to the applied one, sugesting that exists there a voltage of foucault and such voltage went to the EGS, to the terminals 5 and 9, altering the oscilation produced by the terminals 6 and 10, which corresponds to the 2 exits (HO) of the drivers IR2113, burning the mosfets.
I made a test with DC current coming from battery and the voltage resulting from the superior part of the board was some milivolts, different from the DC current arising from the switched source, that presents a little oscilation.
Have you had any problems of that type? If yes, what did you do to solve this problem?
What is the of the thickness of the power board? I believe that must be of 2 and 3mm, and I believe that increasing the dieletric may solve the problem, but this is just a theory and I would like to know your opinion and if you had done some tests.
Thanks.
Regards.
-
As I said we discarded the egs002 board when we found build and component discrepancies on the boards.
We developed our own control boards with the EG8010 chip from the EGS002 board, but we keep our boards simple and robust. Now we can buy EG8010 chips in batches of 10 or so. NB the EG8010 should be the EG8010 HN1518HT1536 version.
[attach=2]
We use 2oz copper, that's double the normal thickness, on both sides of the Power board. I also allow large solder pads for spreading the contact area evenly.
We keep the gates as close as possible to the FETS, a few mm, as parasitic frequencies can creep in.
Photos of your boards please.
Here is mine............
-
Actually my problem was caused by the eg002 with defect, because one of the arms heated too much, there was an amperage consumption of the order of 11 amps in the void and thus it was difficult.
I have thought the best way would be to replace that eg8010 chip with an atmega with SPWM programming.
-
Nice board.
Please remember that the 8010 chip has a soft start, ie, it ramps up the power to the toroid in a few seconds, rather than going in at full battery voltage.
With the 8010, 23.4kHZ, If I recall, each HZ has 27 steps each side.
As I said, above 2kW the toroid is a ferocious animal and needs to be ramped up carefully. With an EW transformer things are different but the losses can be significant.
There are several builds on thebackshed Forum, Warpspeed is the toroid man, where programming PWM has got good results, but I have yet to see a full 6kW toroid Inverter MegaPWM fully working in the field.
I like the 8010 chip, yes the pin 6 start stop is a pain to get correct, but the 8010 is nice and simple to get working consistently.
-
New design board.
-
Nice looking board, dr. Henrique.
Looking at the PCB, I presume you are still using the 8010 chip?
Do you have a mega in there?
If you are using the 8010, it seems that techy folk have been analysing the stop start process on Pin 6. They think that due to sloppy code writing in the chip there is a possibility of a 'Shoot through Event' on the FETS. ..see 'Poida's' and 'Gaspo' comments ........ http://www.thebackshed.com/forum/forum_posts.asp?TID=10786&PN=1&TPN=3 (http://www.thebackshed.com/forum/forum_posts.asp?TID=10786&PN=1&TPN=3)
With our particular design of the OzInverter we do not see any problems with the start stop, IF YOU STICK TO THE BUILD PROCCESS, that Oztules has outlined. ....... and my comments...... ....
Notes. October 2017 Please see Web Site updates.
This Oz Control board, uses the 8010 chip Pin 6 for starting and stopping the OzInverter safely. However, when the power board is first switched on using the 40r resistor to slowly charge the big capacitors using a push button before switching the main breaker on, then the control board may hang, use the reset button briefly to clear. The Over Temp probe sensor connector will also switch ON/OFF, and is a safe way to start and can also be used for a Low Voltage Disconnect, LVD, circuit. The on/Off on PIN 6 is sensitive and can pick up parasitic signals, so keep cables short, less than 150mm or de-couple with caps 100nf & 10uf..
Originally I wanted a hard working Inverter that would be up to the Job, be as simple as possible, be robust so that environmental, insects, humidity etc. would not be as much as a problem as modern PCB's. And we wanted something that can be easily repaired even with dubious spec of the modern day Chinese components.
I will test 'Gaspo's fix as he has modified the Sub control board that has the 8010 chip. see below pics.
As I have said with the correctly built OzInverter we do not see the problems, but I understand that some folk in this day and age just want to continually switch it on and off like a domestic appliance, so their may be an issue if abused ? But most of us when the OzInverter is switched on it normally stays on 24 / 7 year after year.
-
Oztules Inverter has been well and truly analysed by folk like that are pros from the Toroid world and pros from the Electronics area.
What amazes' me is that after all testing etc, Oztules has definitely hit the 'Sweet Spot' with this particular design, yes a few others have added peripheries, and I have backwards design for that 'Simple. Robust and Cost Effective' concept. But in general his simple design is a masterpiece as machines go. The more i see the results, the anylist, the reports, the more I get impressed with Johns 'Oztules' endevours. A clever man, but he says he is "just lucky"!
-
Smart thoughtful innovative people make their own "luck"...
Rgds
Damon
-
Hi Damon,
When you are next at your nearby neighbours house, Her Maj at Buck Palace, having a nice cup of tea and cake, can you mention 'Oztules' John and Hugh Piggott Please?
A present I think these guys deserve some sort of recognition for their RE developments, publications and their willingness to disseminate their IP & intellectual wisdom.......
-
Ha, if only!
The nearest I got was a chat with Prince Andrew about how many rooms Buck House has and if it could benefit from Radbot during its refurb. Not quite the same as a firm order, or the ability to whisper suggestions for honours! %^P
Rgds
Damon
-
Hello Guys,
Can any one help me get compnents to assemble this power inverter?
-
Hello Guys,
Can any one help me get compnents to assemble this power inverter?
Hi yuval,
Welcome to this forum.
Here is a link to get my 'OzInverter 2nd edition' book and the 3 necessary PCB boards.
I offer the 3 PCB's as an extra option, and can be tucked inside the book to save postage.
In the book I also list all the required components for the PCB's and the OzInverter.
https://levivray.com/new-book--make-a-6kw-inverter.php (https://levivray.com/new-book--make-a-6kw-inverter.php)
The PCB boards ….
All full size PCB masks are given in the book for etching your own boards.
Deliberately designed for working OzInverters, ie to allow for all types of hard environmental conditions, and standard easily obtainable through hole components, large solder pads, large tracks, and importantly easy to repair/modify without getting to much substrate separation. ie, it can be kept going.
The power board …. is double sided and double copper thickness and designed for simple through hole components. Espeacial attentention has been taken in getting the 'gates' as close as possible to the FET's and having sufficient track widths and pad sizes to ensure no pinch points. Please remember that this board may have to handle between 300 to 500 amp surges.
The OzControl board ..... There are 2 designs, one for 60HZ and one for 50HZ. Although its a double sided board it can be etched on one side and links attached on the other, all masks are given in the book. This is based on Oztules simple original design, and has been modified for all the latest additions. Again its been designed for through hole components, except the SMD Chip 8010, which has its own daughter PCB. Design principles have been adopted to avoid closed loops and good sized tracks and pads used to avoid pinch points.
The OzCooling board ..... This PCB follows Oztules original concept and is a double board with two independent circuits. One runs the 2 PCB cooling fans, and the other runs the 2 toroid cooling fans. Fans and cooling are very important so the circuits are simple and we use robust and easily obtainable through hole components.
Notes,
The OzInverter has evolved to be a real hard working machine, BUT the book build instructions must be followed.
Important … If its in the book, it's there for a reason.
I trust this helps.
[attach=1]