 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 : Wiseguy New Inverter Build Nano R6
  Page 9 of 13   |
|||||
| Author | Message | ||||
| rogerdw Guru  Joined: 22/10/2019 Location: AustraliaPosts: 852 |
Wow, fascinating background Mick and awesome location too. 1973 was about when I left school and started my electronics career. I'm with Aaron, I'd love to hear more too. Maybe start a thread in "Other Stuff" for motors and generators perhaps ... dunno what's best ... ... or the Solar section or the Electronics section depending on what you'd rather focus on. What are you working on specifically right now? It can be a bit overwhelming but there's no question there's some amazing talent here, with lots of support and encouragement which can make a huge difference in our progress. Cheers, Roger |
||||
wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
Great reading of your background thanks for the information. While most of us have been looking for a source of dead inverters, it sounds like you have an ongoing stream of them in your own back yard, talk about lucky lol. I am surprised about the number of inverters that have been killed in your installation from just being in a similar environment to a lot of us. Yet either others don't volunteer the information or hopefully they have been having better luck than you. What do you think is different about your install that sounds like the panels are arranged in a target formation (sorry to sound flippant) but there must be something peculiar to your location/install ? I am thinking high up and very exposed. I understand perfectly well that lightning activity and semiconductors direct connections to mains wiring (otherwise known as trans-former-less GTI's usually with a TL somewhere in the model number) are bad news. I would have thought that a toroidal transformer type inverter would survive much better given your experience. Dont get me wrong if there is a direct hit to your mains supply all bets are off as to whether anything survives. Are there serious overvoltage surge arrestors installed on the mains side ? Are the runs from the cables to the inverters overly long ? I look forward to some further insight. I think that's enough questions to a two finger typist (edit I actually was using one finger when I started ~ 1985) a smarty walking past said "cant you use two fingers" my immediate two finger rude hand gesture showed him I sure could. Edited 2024-07-21 12:33 by wiseguy If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
Revlac Guru Joined: 31/12/2016 Location: AustraliaPosts: 1021 |
I would say the type of inverter (TL or proper transformer) makes ZERO difference when there is a direct or very close online strike, being grid tied or beside the grid as a backup if you need it, will give you the greatest chance of destruction, being full Off Grid will save you from all of the power grid strikes.  @Mick I started a thread long ago,Off Grid and lightning I need to update that soon as a lot has happened to neighbours And I have been running trouble free the all that time. Short and sweet is to be off grid with your own inverter and have some immunity from trouble the grid will give you. Edited 2024-07-21 18:31 by Revlac Cheers Aaron Off The Grid |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
Hi Aaron, I did speak from some personal experience re the transformers giving the electronics a better chance of survival in electrical storm conditions. I also said above that for a direct lightning strike all bets are off. I designed and manufactured a robust rail crossing charger for a Rail authority in SA who shall remain nameless and the same unit is also used in Tasmania, there were well over 400 sites installed in total, in use to this day. Prior to this they had a brief experience with offline switch mode chargers that failed regularly and these were not direct lightning strikes, even a strike near to the supply wiring could induce enough energy to take them out. During the use period of these chargers, if there was electrical activity at Port Augusta or similar open and remote location it was almost a given that the charger would fail often with no charred or other evidence of lightning strike but dead FETs and fuse & no go. Galvanic isolation provided by the mains transformer design reduced these failures to essentially zero. From my experience I would not hesitate to isolate the mains with a transformer if that was determined to be the likely cause of killing my inverters. I also totally agree that being offline will definitely remove the mains as being the culprit. :) Edited 2024-07-21 19:26 by wiseguy If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| Revlac Guru Joined: 31/12/2016 Location: AustraliaPosts: 1021 |
Hi Mike, Thanks for setting me straight, whats your thoughts on isolating the PV input to the solar system as well? I remove my solar input when there is a nasty storm approaching, some Anderson connectors are just pulled, I think its better to have done what you can, Only other problem is I cant remove it if I'm not home. Getting back to the inverter, whats the minimum size heat sink you would suggest, I guess something at least twice the size of the fet board? Cheers Aaron Off The Grid |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
If I am aware of an approaching electrical storm or if travelling away for a few weeks I still turn off power points for most appliances especially big screen TVs etc to avoid a disaster. You raise an interesting point though, we have 13.5kW on the roof and I never considered disconnecting the PV input or mains to the inverter. We do have some hefty surge arrestors in the switchboard and I do get some comfort in knowing that there are taller structures (yes some are power poles  ) and even taller trees around us.I think if I was in a more remote location and relied maybe 100% on my off grid solar power your suggestion to temporarily isolate the panels from the MPPTs/GTIs for impending storms is a good idea. When I was in my early teens we had 2 decommissioned 32V Dunlite towers and I wanted a good antenna for shortwave so I fitted a porcelain insulator each end and ran a wire between them with a feed down to my workshop. One warm no clouds light breeze balmy afternoon I entered the workshop to investigate a weird buzzing sound, there were blue sparks in a few places on the wooden bench top and when traced out it came from the disconnected antenna feed and was discharging via odd bits and pieces to a ground point I could easily throw multiple continuous sparks 5-10mm long to ground. Needless to say when I realised the implications of lightning activity I quickly removed and dismantled the antenna which had seemed like a good idea at the time. The point of the story is that if steps have not been taken to ground the roof or and at least the rails for the Panels, you dont need a lightning strike, just atmospheric ionisation could cause a static build up and flashover. This is really only relevant to a handy man home install, if its setup on a wooden frame for instance, maybe grounding the panel frames might not have been considered. Re the heatsink I would be looking for something around the same size as the power PCB, the old Aerosharp 1.5kW is fine. Given the quantities of the GTIs they made, it always intrigued me why they chose horizontal fins for their heatsinks, when vertical fins would actually achieve better needed airflow, they certainly had the resources to specify something that would have worked much better for the same price but never did. Edited 2024-07-22 15:09 by wiseguy If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| Eddy C Newbie  Joined: 17/07/2024 Location: AustraliaPosts: 5 |
wiseguy, despite not seeing I cant see having sent PM??? Mick |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
Message was received ok thanks Mick, will reply via PM. If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| Ziki_the Newbie Joined: 13/04/2023 Location: YugoslaviaPosts: 39 |
Hello everyone, I am slowly collecting components, looking at what I have and what to order. I found some igbt half-bridge modules. For a very decent current. They are Mitsubishi CM600DU-24FA, could I use them for inverter or vairiac? Would it be enough to change the gate drive voltage (12->15V)? Maybe a gate driver ic... I'm open to suggestions, no rush.. Thanks everyone. Pozdrav iz Srbije |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
Hi Ziki, those devices are really meant for High voltage switching at high current. Whilst they could certainly handle the current required, their losses due to their inherent higher VCE when turned on would make them quite inefficient. For instance 6kW would be an input current of ~ 130A from 48V, these devices would be losing ~ 500W of power as heat, nearly 20% loss due to their ~2V C-E saturation for each switch. If your battery source was 200V instead of 50V then the losses would be ~120W. I have not tried IGBT's on the inverter. Whilst I encourage people to play my suggestion is to just build the inverter as presented, when it is working, by all means experiment as you then also have a working reference unit to compare to. If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| Ziki_the Newbie Joined: 13/04/2023 Location: YugoslaviaPosts: 39 |
Mike, Thanks for the detailed explanation. the components remain for some next project. When you look at the numbers, everything is clear.  Thanks for the help once again..Pozdrav iz Srbije |
||||
| Eddy C Newbie Joined: 17/07/2024 Location: AustraliaPosts: 5 |
I actually wanted to ask a Question:- 1. With multiple power supplies incorporated into driving the Nano controller and inverter FET PCB, has anyone simulated individual voltage module failures in full flight for impact? (note to self before forgetting; ordering next PCB’s from JLCPCB.com include/combine the parts order, HY5608W FET’s and wiseguy BOM list bits from LCSC. PCB Bds + Parts Combined was free shipping in Feb24.) 2. The FET PCB = 260mm x 108mm. The 2 x Capacitor boards appear to extend, overhang. What is the clearance length required for the FET PCB with Capacitor PCB's mounted? My Oztules 5kw FET PCB is now ready to test with 4 x FET’s fitted. It's only 180mm long x 171mm wide + fans. I’m preparing a Ply sheet to screw a complete set of PCB’s on for testing. Ensuring leaving sufficient room for the wiseguy FET PCB with Caps mounted. Regards Mick |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
I believe KeepIS had tried (all?) various fault modes, there is the main battery supply (typically ~ 50VDC) that converts down to 12V which is not used by the controller card except for powering the on board logic 5V regulator and for powering any relays that may be connected to the relay FET switches. The 12V is also passed along to the FET Power board to power the 3 x isolated 12VDC-12VDC power supplies. If the 5V dies the logic and opto drives stop. if the 12V dies the Power PCB won't function - if the 12V deteriorates to a lower value then the LV cutoff of the opto FET drives will stop them from driving the FETs. I contacted JLCPCB who deny any knowledge that this actually happens, they said "you need to contact LCSC to ask whether they are able to combine their shipment with the same carrier that JLCPCB uses" - I haven't gone any further with this. I am not saying that others are incorrect in their experience, but given the companies seem to have a connection maybe their left hand did not know what their right hand was up to and it may or may not still be this way ??? Edit: I just went to the JLCPCB site and in the left column there is a Parts Manager button, which opens up further buttons, one is order parts so I pressed it and tried HY5608 their price was $1.93 compared to LCSC $1.81 I believe both is $USD. I have no idea if JLPCB stock lines are exactly what LCSC uses but this is a way to order some parts and avoid extra the shipping costs from LCSC which from memory are about $17USD minimum) per order(. Later Edit: Well what do you know, JLCPCB use a part number of C2760540 for the HY5608, which is the exact number LCSC uses, what an amazing coincidence? I expect if JLPCB boards and JLPCB parts are ordered at the same time shipping will be combined Even Later Edit: Yesterday I placed an order for 5 x PCBs today I just went to confirm LCSC's shipping price for 30 HY5608s and it shows as Shipping Free which I guess confirms what was said by others and still holds - place order for PCBs and a few hours later place an order with LCSC, you get LCSC's best pricing and free shipping !! The capacitor boards do not overhang the Power PCB. However if they are mounted with any screw slackness bias maybe ~0.25mm overhang might be possible. The delivered Capacitor PCBs would have an overhang if the excess area was not cut off. The overhang part is meant to be removed (cut along the line) and then further cut down along the silk screen lines to form 20 little predrilled square insulators which can be used on top of the TO220 insulating bushes to further protect any screw/nut/spring washer damage to the bushes whilst also providing extra clearance. If the largest diameter capacitors are used ~35mm then they overhand the edge of the CAP board to the outside by about 2mm. Edited 2024-07-29 12:03 by wiseguy If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| Eddy C Newbie Joined: 17/07/2024 Location: AustraliaPosts: 5 |
Thank you Mike, I was tossing up where to ask. Here or KeepIS. I appreciate your time is currently tight and I’m not in a rush. Two weeks already passes too fast for me. While posting remembered the combined freight option and left the note to insure being found before I forget. I have not been off the block to purchase a diary this year. I too will also be after 100x HY608W plus the other bit listed in the board BOM’s. Having components mounted also attracted free post? Something in return for using the slowest post option? Not signing in I to forget to set the country and US$ to AU$. Also had to be careful confirming links remained under the company umbrella. There were other cloned sites trying to grab your order. Always having to check the hpps address remained correct. Ply test and repair board. Ready to test and doing the ply board test and repair station first for once. Once an inverter is up and going in its box, like manuals, left to last and never get Rountuit. Shuffling bits around, arranging for safe access and measuring without reaching over 240v bits. The Oz control needs a hard cover over the Vfb transformer and its 240v connection. One 240volt bite as a kid was enough. Plenty of RF burns; there like pushing a pin into your finger tip and hitting the bone. Teaches you to keep hands in pockets and not point in transmitter rooms. Ensuring leaving room for your FET board noticed images KeepIS had posted would exceed the 260mm x 108mm.  I had read the insulator spaces were removable and would be using. (Well done, good foresight planning ahead) I still have 6 months lead time to complete my in depth drill down of your boards. Comprehending variations in design bringing the FET’s to a safe start /stop sparks interest. Regards Mick |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
KeepIS had special capacitor boards made for his build as he was using large screw mount capacitors. I was referring to the boards that would be supplied through me as part of the inverter project build that I presented here. Re using JLCPCB to populate boards - personally I would use them if it was a surface mount board with many SMD components. For the through hole PCBs I try to maintain ability for home builds with many components being sourced from our parts bins. Also the many non mainstream parts (ie high current terminals) used in the inverter and non standard lead bends of the power Mosfets etc could be a bit of a challenge for them. Also not sure what documentation and pictures they would require for a project like this. I go to a lot of trouble to ensure accurately aligned Mosfets to the mounting holes on the power PCB by using a long necked TO220 mounting bush and 3mm screws etc and ensure the 3 legs are pretty central to their pads, this is repeated for all 16 FETs. It means if I ever need to remove a MOSFET half the battle is already won if the legs are not under tension against the internal bore of the flow through joint, a few seconds with the solder sucker and they just fall out. I am confident that this sort of attention to detail would not occur with their builds, the low cost would not allow it. Edited 2024-07-30 02:25 by wiseguy If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1679 |
Hi, those pictures show capacitor test boards that have not had the insulator section cut off, and as Mike pointed out, these special Cap boards are a few mm wider and do overhang slightly even with the fiberglass insulator pads cut off. On reflection, I should not have used 10,000uf caps, I think 4,700uf caps with 2 or 4 per board would be plenty of capacitance and cost a lot less. But that's just my opinion after a lot of testing in my inverters. Nice to see more of us building the inverter  It's all too hard. Mike. |
||||
Bryan1 Guru Joined: 22/02/2006 Location: AustraliaPosts: 1343 |
Well had a bit of time this arvo so decided to make a start on the inverter control board and I do have most of the resistors here. So the plan is to use what I have here and with the missing ones get onto digikey to get the rest of the components for both this board and the one's I'm missing on the Mad board. Now as I do have 25 of those HY5608 fets here by going 3 each leg on the Mad board left me one short on this inverter build. So got onto LCSC and got 50 fets coming also got a few parts for this inverter build. I did find trying to search parts on LCSC wasn't the easiest so just left it for the digikey order. Regards Bryan |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
It was time to actually see what peak currents are involved in a 3.5kW toroid when it is dropped across the mains asynchronously. I used the new second stage switcher to provide the degauss before closing the relay that bypasses the soft start in rush current NTC limiter, connecting the toroid directly to the mains. The purpose of this is that the mains will be provided by the inverter stage1 and I thought it prudent to confirm what peak currents to expect, so the correct value of inrush NTC resistor can be chosen with something other than an educated guess. I also wanted to see the current waveform for a few successive cycles as the current peaks taper off as we transition away from saturation. I could only find 5R NTC inrush limiters in my parts bin so I fitted them in the two locations giving me 10R of limiting. I will provide some CRO pictures of the current peaks I encountered in my next post but I have run out of time atm. The 10R resistors allowed me to capture approx 25A peaks worst case so it would appear that when we find saturation at switch on, the transformer is a very very low impedance as expected. The current peak is a one off event for that half cycle and then quickly reduces to around 20% on the second half cycle and after about 500m Seconds has settled to its low running value. If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1679 |
Hi Mike, I have a few 10R limiters here so very interested in what you find. Just so I'm on the same page, where are the current peaks being measured. Unless I get interrupted, I'm looking at having this Inverter running sometime Monday, and if all goes well, I'll be adding the 2 stage switcher then. _ It's all too hard. Mike. |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1679 |
I'm thinking a bit clearer this morning, switching the second toroid across the AC output is no different than switching a huge Toriod Variac cross the existing inverter, very big current surge and half the time tripping the inverter with over 600A Peak DC input surge. I'm assuming that having around 20R in series as a pre-degauss substantially reduces the inrush current on stage one. When starting the inverter wired as a full dual power stage inverter, it should be no different than starting it as a single stage inverter, as both toriods would be soft started together from zero. _ Edited 2024-08-17 08:58 by KeepIS It's all too hard. Mike. |
||||
| Page 9 of 13 |
|||||
 Print this page |
