 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. |
Forum Index : Electronics : 6Kw Ozinverter build
  Page 45 of 51   |
|||||
| Author | Message | ||||
Madness Guru  Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
Still in holiday mode, just got around to washing the car this afternoon. Have not thought about working on that Inverter, I do have all the Electronics ready though. Weekend is booked out too have birthday party tomorrow and catchup with friends on Sunday. Here is a series of 3 videos about one place I visited. About an Australian bloke called John Spies and what he has been up to for over 30 years in the Golden Triangle and the place they have built there called Cave Lodge. It is quite a story. 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 |
Mark, I see you need further explanations about what happens at the rds on tests we do. That probe is measuring *millivolts* which, by passing exactly ten Amps through the Mosfet legs and shifting the millivolt meter decimal point one place, you can then interpret the reading directly as milli Ohms. The current through the legs does not change, no matter how long they are. Its just that slightly longer legs have a miniscule higher voltage drop which your sensitive digital millivolt meter can easily detect. So, to get consistend rds on readings you must measure each mosfet at exactly the same position on its leg. That is easiest right against the mosfet body. If you do some research you will find solder is not exactly a brilliant conductor when compared to copper. Why do you think proper crimped terminations are more reliable than soldered ones? Consider a well engineered screw terminal like a crimped connection. I'm sure you are not familiar with the terminal I use - it does *NOT* jam a screw into the mosfet leg to make the connection. It is rather a movable housing that clamps the leg against a fixed clamping plate as the screw is tightened. If the mosfet leg is examined after it having been clamped you will see several parallel grooves where the clamping surfaces had dug in. I would say its very close like crimping. How long would a home build inverter last? How long is a piece of string?  .I'm confident mine would outlast yours by a long way after I finished tweaking them to my satisfaction.  With your soldered mosfet PCB I would ditch the board and assemble a replacement if more than one mosfet had blown. Its just too easy to wreck plated through holes by extracting semi vaporised mosfet and other component legs. I learned that the hard way and I gather you too  .Klaus |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
That is what mostly put me off soldered in high power mosfets, and soldered very large wires directly onto circuit boards. Large components solder in very easily, but unsoldering requires much more heat, and it tends to destroy any plated through holes. If the fault current and resulting explosion did not burn up the tracks and blacken the board, repeated unsoldering surely will cause tracks to lift. Once your circuit board tracks become very severely damaged, or destroyed, its not really possible to make a nice clean neat repair of it, and its unlikely to ever be as robust or reliable as it originally was. If you really prefer soldering in the mosfets, and there are some good arguments for that, at least do it on a small disposable sub board that then becomes a throw away disposable item each blow up. Soldering a huge high current wire directly onto a circuit board copper plane has similar problems. Much better to bolt a lug onto the board with a pair of large flat washers to spread out the point of high current density (and potential heating and voltage drop) where it feeds into the copper layer. Its also mechanically a lot stronger with the big washers. Many clever ways and means of designing high powered equipment with circuit boards, so that its both more reliable and very easy to make a proper repair, and repair repeatedly without the whole thing becoming total junk after about the second really big blow up. You will also find that if you think all of these finer points of construction right through, and maybe come up with a few ideas of your own, things become much less likely to blow up in the first place. Cheers, Tony. |
||||
renewableMark Guru Joined: 09/12/2017 Location: AustraliaPosts: 1678 |
Klaus what high power sockets are there that can take loads like that? My concern was the amps they would need to have going through them and if they came loose, also running them with long skinny legs (but I suppose you could trim the legs down). Cheers Caveman Mark Off grid eastern Melb |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
Look at the current flow cross sectional area of the pins on the mosfet. Look at the physical bulk of the brass in the screw terminal block. The pins on the mosfet are vastly thinner and more frail than the terminal block. If anything is going to burn up first, its going to be the legs or silicon die in the mosfet, and you cannot do very much about that. Its always going to run hottest and fail at the weakest point, and that will be the mosfet itself. Cheers, Tony. |
||||
| renewableMark Guru Joined: 09/12/2017 Location: AustraliaPosts: 1678 |
LOL when you said golden triangle I thought HUH that's here in Vic here , but they obviously have one too. I need to get a wriggle on and finish the backup unit so I can cut the grid connection completely. No point having it there if it's never used, $101 a quarter just for connection. So far we have been going to float anywhere from 9am to noon on a cloudy day. Still haven't put all the panels up yet, can fit another 2500 west and 1500 north and 1500 south(for cloudy days) We have 8000w east west up so far and it's coping really well. Since I got to that amount of panels on the roof we haven't used mains since late Aug, almost 2 months. Cheers Caveman Mark Off grid eastern Melb |
||||
| renewableMark Guru Joined: 09/12/2017 Location: AustraliaPosts: 1678 |
Thanks Tony have you got a link to them? Cheers Caveman Mark Off grid eastern Melb |
||||
| brucedownunder2 Guru Joined: 14/09/2005 Location: AustraliaPosts: 1548 |
You are doing fine Mark, sounds like you have a good set up down there . Any chance of a few pics of your controller/switching panel /battery /inverter arrangement ?? I might like to steal some better ideas from your installation . I've plenty of pics, but you'll have to search ,"brucedownunder' and pour yourself a beer while searching ,it'll take a while ,I've been here for years. all the best , follow your posts every day Bruce Bushboy |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
Mark, It depends which type you need. There are two sizes 0.2 inch pitch spacing for the larger TO247 "big" mosfets, and 0.15 inch spacing for the TO220 "small" mosfets. Also they come in both horizontal and vertical types. Horizontal have screws on the top, and the mosfet sits parallel with the circuit board. Vertical have screws on the side, and the mosfet stands straight up, which is probably the type you need for an Oz type inverter. Altronics have all four types in their catalogue. The ones I think you need are for the big vertical mosfets. Altronics part number is P2065, these are three way, vertical, 5mm pitch and $1.40 each.  Manufactured by Dinkle and are their EK500R series very conservatively rated for 20 amps continuous.https://www.electronicsdatasheets.com/manufacturers/dinkle/parts/ek500r02p When you hot foot it around to Altronics, take an HY4008 with you just to be absolutely certain its the right part. These all start to look very similar when you are looking over dozens of different types and styles, and its confusing to judge by eye. Last time I visited Altronics looking for these they were out of stock, as the vertical type are far less commonly available and they probably do not sell a lot of them. I have never seen the vertical type advertised on e-bay, although the horizontal type are readily available from many different sellers. If they ever do come up I will buy a zillion of them. Anyhow, ordered a decent quantity from Altronics and paid, and they arrived by mail in a few days (from Perth). Cheers, Tony. |
||||
| Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
If I need to remove MOSFETs I bend them backwards and forward a few times and they just snap off, then it is very easy to heat each pin individually and remove them. The bend happens above the solder and there is no damage to the PCB, neither is there any damage due to the PCB when removing each pin individually as you only need to apply minimal heat till each pin comes out. Any comments about solder joints failing is invalid as you have to solder in the screw terminals also. I fail to see how puny terminals rated at 20 amps is better, plus if you learn how to solder you don't need to damage the PCB. Might be good for to doing breadboard type testing where lots failures are to be expected. Today I showed this thread to an uncle of mine who is a highly qualified Electro-Mechanical Engineer and now a lecturer at a University, he was rather amused by this concept. There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
| Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
Never heard of a Victorian Golden Triangle, I am sure the one in Thailand, Laos and Burma is much better-known worldwide. There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
| renewableMark Guru Joined: 09/12/2017 Location: AustraliaPosts: 1678 |
Prob 90% of Aust's gold has come out of there, it has the Ballarat/Bendigo fields. I gave up detecting, been too flogged now, just use the dryblower and highbanker these days. Good fun on a nice day with some beers. Cheers Caveman Mark Off grid eastern Melb |
||||
| renewableMark Guru Joined: 09/12/2017 Location: AustraliaPosts: 1678 |
Bruce, I'll take some pics for you mate. The current inverter has it's covers on now, but will be assembling the backup unit very shortly. I have the Mad solar panel controller operating with just the ELV 2 in series panel side going. I do have a couple of Latronics grid tied inverters that I'll run through that too on the other side of the board, but need to put those panels and wiring up yet. Plenty of materials lying around, just need the time and inspiration to do it. Wife, kid, business and elbow bending gets in the way a bit. BTW Bruce we did go for a little holiday to Echuca, took the van and stayed on the banks of the Murray at Barmah state forest, it's a lovely area, saw pleny of roo's and a emu family with 30cm high chics came close by each day. Tony, re screw terminals, yeah I'm not sure about them, 20A seems a bit low. I'm not totally stubborn and will use different options if they are better, but just not convinced with them. I have those terminals on a solar controller in the caravan and one in the 24v backup system in the shed and it seems like you can torque up the screws just a whisker every time you look at it (every few months I check them) Anyway we don't have to all do everything the same. Cheers Caveman Mark Off grid eastern Melb |
||||
| renewableMark Guru Joined: 09/12/2017 Location: AustraliaPosts: 1678 |
Yes correct, but when you seat a fet down on the board the thicker part of the fet leg comes into contact with the board, and if enough solder is applied there is a greater contact area compared to the lower skinny section of the fet legs. Anyway I don't think it's a really big deal, more of a preference thing. Time for more bending of my elbow, cheers fellas. Cheers Caveman Mark Off grid eastern Melb |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
You are obviously not an experienced service tech or equipment designer. So you have (say) four or five parallel mosfets soldered into a board with a source to drain dead short. Maybe they are all shorted, or maybe only one. 1/ You can snap them all off and then test them to find which particular one is shorted, clear away all the solder and replace all five, because the good ones now have legs too short to replace. 2/ Quickly unscrew all mosfets, test them, and put back the original good ones. Should only take a couple of minutes tops to do that. Fast, easy, and no damage. The safest thing to do after any blow up is to remove and test all the mosfets, otherwise if you miss one, its only going to blow up again. If you have to snap off all the mosfets, and replace them all, that can be a lot of work and very wasteful of mosfets. You may fail to see it because you do not think like an engineer. With sine wave PWM the avearge duty cycle for each mosfet will be 50%. In other words throughout the cycle the duty cycle varies between something more than 0% to something less than 100%. At the zero crossings it will always be 50%. If the sinewave has 40 amps rms per mosfet, each mosfet is on for only 50% of the time and sees only 20 amps AVERAGE current flow. If you have three mosfets per leg you can do 120 amps rms without exceeding 20 amps average per mosfet. At 48v that is more than 5.7Kw. As many here are running four mosfets per leg at 5Kw, those terminal blocks are going to be just fine. So it IS perfectly sound engineering even if you yourself fail to see how it could possibly work. Glad to hear he has a sense of humor. If you don't like the screw terminal block idea then don't do it. But others on the Forum may like to try this, and some that have actually done it, are very pleased with the outcome. Nobody has yet come back and said it did not work or has caused a problem. Cheers, Tony. |
||||
| Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
When there is a failure it usually takes out all the FETs in that leg of the H Bridge, not just one or two, how do I know that? from EXPERIENCE. Anyone who has played with a serious version of these inverters would be more than aware of that. As for the occasional survivor, why would you risk putting a stressed component back in with brand new good ones? There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
| Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
Okay I am with you now, we have Gympie just up the road. Not as big as Victoria's gold rush but big enough to save Queensland from bankruptcy. There is supposed to be a stash of gold hidden in Northern Thailand at the end of WWII by Japanese soldiers escaping from Burma. Their truck broke down and they commandeered 3 elephants from local hill tribe people but the Elephants were unable to carry the amount they had. 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 |
That is all very true, but I have sometimes had situations where only a single mosfet has died. But if its a big high power blow up usually the lot are definitely all stone dead. Anyhow, the screw terminal block idea is proven to work, and certainly greatly reduces the amount of time and effort required do a full repair which has to be worth something. IC sockets are pretty good things to have as well for all of the same reasons. Occasionally a screw terminal block is blackened or burned, or even partly melted during a really big blow up. Then of course that needs to be replaced as well. But much more often the screw terminal block escapes completely undamaged. A screw terminal block definitely keeps the exploding legs well clear from the circuit board, and just about guaranteed the board itself will not be blackened. A further idea along those lines is to cover the mosfet legs with a short length of reasonably thick wall teflon tubing to contain the exploding legs. That might protect the screw terminal block from damage to some extent. I am now doing this, but its been almost a year since I last blew a mosfet, so I am still waiting to see how effective this teflon sleeve idea is in practice. Or maybe you think that is a really crappy idea too ? Cheers, Tony. |
||||
| johnmc Senior Member  Joined: 21/01/2011 Location: AustraliaPosts: 282 |
Good day all. Mosfet replacement is not the problem, it is the collateral damage to the control circuit,s. Mosfets are cheap and it is not worth the time or effort trying to salvage a stressed mosfet . I feel that the isolated mosfet power board is the only answer. cheer john johnmc |
||||
| Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
Using an 80A DC breaker for the DC input damage is limited to the internals of the MOSFETs. There are no burnt legs, no exploding cases, no burnt PCB and no damage to any other components. So repair is as easy as snap off the MOSFET legs and desolder the legs, solder in new MOSFETs, clean off the flux and back in business. But it has been a long time since I have needed to replace any FETs. HY4008Ws cost $1.49 each from that supplier (never had a bad one from them) so I don't worry about the cost of throwing any in the bin that I have any doubt about. There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
| Page 45 of 51 |
|||||
 Print this page |
