Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.
johnmc Senior Member Joined: 21/01/2011 Location: AustraliaPosts: 282
Posted: 12:36am 13 Nov 2017
Copy link to clipboard
Print this post
Most Impressive workmanship Klaus. Thanks for the effort, to show and tell your Inverter build. john johnmc
Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498
Posted: 12:43am 13 Nov 2017
Copy link to clipboard
Print this post
I am always amazed at how everything you build Klaus looks like you bought it for a lot of money.
I can understand you putting wheels under it, I cheated and assembled my Inverter on the wall. Too bloody heavy to lift, I had a hernia fixed a few years ago, don't want to go through that again.There are only 10 types of people in the world: those who understand binary, and those who don't.
brac321 Regular Member Joined: 30/11/2016 Location: SloveniaPosts: 54
Posted: 01:19am 13 Nov 2017
Copy link to clipboard
Print this post
Nice design, Klaus!OffGrid 24/7: 250w black mono solar panels 6.4kWp, 1x Midnite Classic 150, 1x MorningStar MPPT 60, battery bank 840 Ah @ 48v, modified LF-8K inverter, DC/DC, hot water harvesting, etc.
Tinker Guru Joined: 07/11/2007 Location: AustraliaPosts: 1904
Posted: 01:47am 13 Nov 2017
Copy link to clipboard
Print this post
Thanks guys, your praise is appreciated.
Mad, thinking about how many mosfets went up in smoke and about other things I bought for it but then did not use, you could say it did end up costing me a good sum, buying it bit by bit
My mates keep telling me that I don't do things by half, whatever that means .
But I enjoyed the journey, making something from scratch that in the end actually works is very satisfying and I don't care about the cost. It certainly kept me on my toes, overcoming the many challenges along the way. Klaus
Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498
Posted: 09:21am 13 Nov 2017
Copy link to clipboard
Print this post
I hear you Klaus, I don't want to think about what I spent. I plan to build a second big inverter shortly as well as a backup unit. It will be very straightforward though now I have found what works.
As Oztules says it is a disease.There are only 10 types of people in the world: those who understand binary, and those who don't.
Tinker Guru Joined: 07/11/2007 Location: AustraliaPosts: 1904
Posted: 12:24pm 14 Nov 2017
Copy link to clipboard
Print this post
Yeah, especially when you had to buy multiple PCB's and am now wondering what to do with the extra ones .
I have already cut, drilled & tapped my spare Aerosharp heatsink and mounted the 3 sections onto plastic brackets. The end fan bracket is also done.
The control PCB's are finished, I had assembled another set for the inverter testing stage. Biggest job is re wiring the spare 3KW toroid I have lying about here, possibly start on that once the weather gets cooler again here.
The soak testing today of my big inverter was interesting. It sat on 31 degrees heatsink temp. after running on idle all night. Solar input was 2.2KW and my battery bank full so I connected every quiet load I could find (strip heaters, flood lights) which totaled to 2750W and watched the heat sink temperature raise.
This was happening very slowly, I wanted to hear when the heatsink fan (EG8010 controlled) came on. After about 30 min the temp. had risen to 41 degrees and I could hear the fan turn on. I then set the independent control of the other fan to also run.
The heat sink temperature continued to creep up but much slower, after about 90 minutes by battery bank had dropped to 90% (2.2KW in and 2.75KW out), it drained slowly. The heatsink had gone up to over 42 degrees by that time.
I reduced load power then to see when there was some equilibrium reached. Near enough 2100W constant load the heat sink temp. stopped raising and fell ever so slowly.
Its unlikely that I will have a constant load of over 2.1KW connected for hours on end, most things here (fridges, machinery) are intermittent load only.
So, performance wise, she's a little ripper . Ambient temperature, BTW, was 25 degrees in my shed.
After this a quick shock test followed, 2750W constant load and then my abrasive metal cut saw which rates at 2200W and has no soft start.
Watching the sine wave on the CRO while turning the saw on showed little sinewavy wiggles on the top and bottom half of the trace for the blink of an eye, then the trace steadied to a nice sine wave.
Now the inverter is powering my house. It is certainly a lot quieter than the modified PJ that was doing this job before. That one had the fans coming on and off all the time on a 25 degree day. But then, these things have a ridiculously small heat sink for something rated at 8KW! Klaus
nickskethisniks Guru Joined: 17/10/2017 Location: BelgiumPosts: 458
Posted: 09:58pm 14 Nov 2017
Copy link to clipboard
Print this post
Wow, nice job!
Is there a rule of thumb to chose the value of the input capacitors?
Tinker Guru Joined: 07/11/2007 Location: AustraliaPosts: 1904
Posted: 10:51am 15 Nov 2017
Copy link to clipboard
Print this post
Thanks.
There might be a rule of thumb, oztules seems to know all of them .
Capacitor voltage *is* important. So is the quality. The ones I choose (10,000uF/100V screw terminal) are expensive but when tested they had no leakage at 70VDC. On the other hand, the snap in type 10,000uF/100v caps I have here started leakage at 65V. It seems you get what you pay for with capacitors too.
The capacity question, depends on the inverter rating. For mine, 6 x 10,000uF seems to be what others used. For a <3Kw job perhaps 4 x 10,000uf. Best err on the generous side as the capacitors make or break the peak current surges expected.
Very important is the location of the capacitors with respect to the Mosfet's.
Also, the cooler they run the longer their life span. Putting them in the hot alley between the heatsink fins is *not* the best location. But it does make the most compact PCB layout.Klaus
oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686
Posted: 10:55pm 15 Nov 2017
Copy link to clipboard
Print this post
I think a few things need to be cleared up here.
Electrolytic caps are interesting devices, that are capable of repairing themselves very well given the chance.
The cheap caps you can buy with their horrible leakage, is generally more to do with the person measuring them than them.
There is a fair chance that they came from China, Hong Kong Vietnam, Taiwan etc etc.
There is also a chance they did not assemble them in a climate controlled environment.
So firstly there will be some hygroscopic dissolution into the dielectric and separators. This will actively attack the oxide layer, reducing the capacitance.
They will also have been in storage without any charge in them.
The outcome will be lower than expected capacitance, and poor leakage characteristics that Tinker discovered in his tests... no surprise so far.
However, if run for say 12 hours, the capacitance will recover to spec, and the leakage will fall after 10 mins or so.
Self healing of the oxide layer is part and parcel of electro's... so don't be alarmed unless you have thoroughly tested them *after* allowing them to come back from the dead... it is normal.
Rule of thumb.... no not really. The best test is" are they getting warm/hot."
This seems to be effected by the number of caps in particular, and their size. Bigger is better. The ripple which we are trying to get rid of, will appear as heat in the poorer esr caps, and still as heat in the better low esr caps.
Surface area to radiate this is the key here. If they are not getting hot when the inverter is under load, you have enough caps.
I found that three caps under duress, ran hot to concerning hot, and with a fourth, they all ran cool.. so the extra capacitance helped far more than it should have.... spread the pain.
The caps do more than just supplying the surge, they need to present the DC source as a low impedance, and this will help quench the spikes delivered from the fets ( it is switching an inductor after all).
So cheap cool caps will last as long or longer than good hot caps.
I do like the capacitors in between the heat sinks... they will stay cooler there than anywhere else.... if you design it to do so.
I make sure the fan comes on at 30c... but due to my circuits, it is really idling plus a bit. It is impossible to hear it, but it is still circulating air across the caps and heat sinks.
My heat sinks rarely get over 40c. I am in a cooler climate yes, but thats what is max normally any day. This will possible be achieved during the 2-3 hours of hot water and washing machine ( no I can't train her to do them at different times.... not that brave).
This means that 60-80 amps of DC will be switched, and so ripple will be present on the caps.. they are air cooled, so their surface area is greatly effectively increased..... this means the crappy caps I use will last for a very long time, as their biggest enemy is kept under control, even under high power long term use.
Caps out in the air without cooling will get warmer than this setup, as the outside air drifting around in the heatsink region will do a very much better job of cooling things, than a box with a transformer running at 50c... my tranny is smaller than Tinkers, so 4kw continuous heats it to 50c after a few hours. When the heatsinks get warmer than 33c the fan drives much harder, and the caps see a cooling stream of external air all the time them.... the heat sinks will stay in the 35-45c as a generality.... the caps will be much cooler than that.
Anything that keeps the caps cooler is worthwhile, and forced air does this best. The caps won't see the heat sink temp, unless there is no fan.
I am not a fan of ( as I once was) of no fan cooling. Provided it is done sensibly and innocuously most of the time, that is sufficient.
I think we were all poisoned by the home computer with a fan screaming its head off all the time. It does not need to be like that. A decent sized fan quietly running in the background is quite comforting really..and makes a huge difference to lifetime.
If I can get 2000hrs@ 105c, then I can get 128000hrs @ 45c or 256000hrs@35c.... 30 years instead of 2000hrs.. Fan cooling with external air is worth the effort if done nicely.
Tinker has 6 nice big caps....I expect the ripple to be low in his application, and the temp to be below 45c most of the time...... they should outlast both of us.
........oztules
Edit... and no....I have no idea why the first part is in double spacingEdited by oztules 2017-11-17Village idiot...or... just another hack out of his depth
Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498
Posted: 11:39pm 15 Nov 2017
Copy link to clipboard
Print this post
I was wonder about the temp with the heatsink also, would they be any better off sitting elsewhere with no fan? As Oz said heatsink temperature is not that high anyway.There are only 10 types of people in the world: those who understand binary, and those who don't.
Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406
Posted: 12:02am 16 Nov 2017
Copy link to clipboard
Print this post
In the old days, fans were noisy, or would become noisy and were often the most unreliable component. That has all improved greatly over the last few decades, and fans rarely cause any trouble these days.
Its also usually fairly easy to use more than one fan for greater airflow and some redundancy. For the "air tunnel" type of construction, one fan pushing and one pulling may not greatly increase the flow, but offers some insurance against a single fan failure.
Oz has pretty much nailed the capacitor question. Low capacitor surface temperature is the key to long life and reliability. Capacitors should not get unduly hot, either the capacitor is rubbish, or you need more of them if they run hot from self heating. Cooling air is a big help, but it is not a good idea to rely completely on forced air to fix a capacitor that is itself generating too much internal heat.Cheers, Tony.
Tinker Guru Joined: 07/11/2007 Location: AustraliaPosts: 1904
Posted: 10:57am 16 Nov 2017
Copy link to clipboard
Print this post
Thanks oztules, you explained that capacitor thing better than anybody I know of.
Yes, mine do run cool also. You might not have spotted it but the fan for heatsink cooling, to which ends its attached, also blows some air on the other side of the PCB, where the capacitors are mounted, So they benefit from that air too. They are also located as far as possible from the big toroid which, BTW, has the exhaust fan mounted directly above, on the toroid side of the enclosure.
Today we had 36 degrees forecast, in my shed it got up to around 35 degrees. My small split reverse cycle aircon was running, plus the two fridges.
The exhaust fan was cycling on and off, I'm still playing with the turn on temperature and hysteresis. I cannot hear the fan on the heatsink inside the enclosure, it ran at a similar temperature when I was testing, but with the covers off and a heat gun tripping the sensor.
I might fit a LED to show when its running and how long - its sensor is on the positive heat sink. The exhaust fan senses temp. on one of the primary heat sinks. As the exhaust fan is smaller - but there is room to fit two - the lot will get closely monitored during the summer. If my temperature display shows more than 40 degrees on hot days the second exhaust fan will certainly go in.
In favor is the air intake, just centimeters above the shed concrete floor, coolest place in there.Klaus
Tinker Guru Joined: 07/11/2007 Location: AustraliaPosts: 1904
Posted: 10:28am 31 Mar 2018
Copy link to clipboard
Print this post
Its time to add another chapter to my long inverter building saga since inverter #3 had just passed its test. No real surprise here as its a very close copy of that monster inverter #2 described on many previous pages of this tread.
This one has just a single 3KW Aerosharp core, rewound to my standards. And adding the extra turns paid off, just 19.3W draw from the battery at stand by. I like that.
The other difference was using only 16 x HY4008 Mosfets and only 4 of the big, expensive, screw terminal capacitors.
The cabinet is a little lower and narrower too as the smaller toroid took up less room.
A picture gallery shown below, taken from all sides, might give other inverter builders some ideas.
Wiring under the top control panel:
Front side:
Left side:
Back side:
Right side:
View from top:
The test results did surprise me. As mentioned, it worked right away and required just adjusting of the AC voltage. Then I started to load it up, here is a pic of the control panel at over 5KW, the trace does look acceptable too. That little brother to my two stack core inverter packs a punch!
Loading continued after this pic was taken to 6200W or so, no problems at all. But I started to worry then about pushing my luck with that 3Kw core plus the jumper lead clamp I used to connect to the battery got a bit warm from the 100+ amps.
I set the over current shut down to trip at 6KW, in use it should never see that load as this inverter is destined for light duty - literally - powering the lights in my place. Of course, it will serve as a reserve in case my bigger inverter, which powers the rest of my house takes a break for some reason.
Inverter #1, my 8 (Chinese)KW PJ conversion is now retired, collecting dust on a shelf.
Here is a pic of my two inverters, #3 is on the left, not yet wired to the old LA battery bank on the shelf above. I will do a heat soak test tomorrow if there is lots of sunshine.
Klaus
Boppa Guru Joined: 08/11/2016 Location: AustraliaPosts: 814
Posted: 05:27pm 31 Mar 2018
Copy link to clipboard
Print this post
You seem to be going the route I am thinking about, low draw inverters for specific tasks One thing I was thinking about, how hard would it be to 'stack' inverters (ie similar to grid ties in a way) where they are shut down until needed (like a low draw one that runs the 'night load', then as it hits its limits, another heavier one cuts in to pick up the slack, maybe another that cuts in when the heavy hitters like driers/aircons fire up
so that idling current is cut back to just the lowest level possible for most times, the heavier ones that draw more 'standby' current are actually switched off until needed
if that makes any sense?
Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498
Posted: 09:26pm 31 Mar 2018
Copy link to clipboard
Print this post
I do something like that with GTI's, so long as the sun is shining. Getting offgrid inverter to synch is something that has not been achieved with these inverters AFAIK.
My big inverter draws 45W at idle and will run 8KW continuously if you have a decent battery bank the idle power is really not worth losing any sleep over. In my case overnight idle power draw is less than 1% of the battery capacity.There are only 10 types of people in the world: those who understand binary, and those who don't.
johnmc Senior Member Joined: 21/01/2011 Location: AustraliaPosts: 282
Posted: 10:33pm 31 Mar 2018
Copy link to clipboard
Print this post
Thanks Tinker for your description of your new #3 build cheers johnjohnmc
Ralph2k6 Senior Member Joined: 24/09/2017 Location: AustraliaPosts: 129
Posted: 11:55am 01 Apr 2018
Copy link to clipboard
Print this post
Some beautiful work there Klaus, certainly very professional looking gear.
I'm hoping to do something similar to what boppa is suggesting as I don't have tine or resources to do the ozinverter any time soon, but I have scored some UPS's (1500VA) cheap at auction which I hope to use for running the house lighting circuits.
Ralph
Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406
Posted: 04:56am 02 Apr 2018
Copy link to clipboard
Print this post
Indeed it is a work of art, and there is a great deal there that many of us here could learn from.Edited by Warpspeed 2018-04-03Cheers, Tony.
Revlac Guru Joined: 31/12/2016 Location: AustraliaPosts: 1024
Posted: 10:06am 02 Apr 2018
Copy link to clipboard
Print this post
I agree, Its a great bit of work there, I'm impressed With the output of that single core. Cheers Aaron Off The Grid
Tinker Guru Joined: 07/11/2007 Location: AustraliaPosts: 1904
Posted: 10:49am 02 Apr 2018
Copy link to clipboard
Print this post
The secret is lots of copper sq mm in the windings IMO. The 16 HY4008 should handle that load easily.
When I did the heat up test, to see when the fans come on (set at 41 degrees) it took absolutely ages with a load of 3.3Kw. That load my battery & panels could maintain during the day without draining the battery bank.
Ambient was around 25 degrees in the shed and once the fans came on the inverter started to cool down. That is where the big cabinet pays off. This time of the year I do not expect the fans to come on at all with the usual cycling loads at my place.Klaus
Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.
Print this post
.
. Ambient temperature, BTW, was 25 degrees in my shed.
