 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 : Inverter building using Wiseguys Power board and the Nano drive board
  Page 10 of 23   |
|||||
| Author | Message | ||||
KeepIS Guru  Joined: 13/10/2014 Location: AustraliaPosts: 1679 |
Once again for anyone interested. Having two Nano controllers, I finally have the luxury of a complete test backup Nano Inverter. I can now test, write and run code in a running loaded inverter without getting out of my chair or unplugging anything or switching anything. So with a spare hour this morning, I completely rewrote the only part of my code that I was not happy with, the Calibration entry code. Changes are included in the version file in the Zip Ver 7.0 EDIT: Just added two minor text message changes that I had overlooked  EDIT: Version 7.1 update: Update removes the need to press "enter" when selecting menu items, and adds a "Command line prompt" when waiting for a voltage, temperature or current value to be entered, obviously followed by a carriage return. FYI All uploaded updates always use the same download link. . Edited 2024-07-01 10:51 by KeepIS It's all too hard. Mike. |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1679 |
Intermittent rain and a bleak day, so I spent the time getting everything accurately calibrated in both inverters, along the way I improved the calibration accuracy when an AC voltage or AC current value is entered in setup. Changes made just before this post and uploaded, still Ver 7.1. I was able to set zero offset in the R7 board with the updated circuity, FYI in the previous R6 board, I selected the best two MCP6002 from a dozen I have in the parts bin, and got close to zero offset in the R6 controller. Testing in both inverters shows that AC current offset is swamped by Solar charger and inverter noise, usually a value of 0.2 to 0.4 on the LCD. Unplug the ACA sensor and zero Volts returns on a running inverter, so noise is coming in through the Current transformer. The test bench unit will show 0.2 with no load and 0.0 with a small resistive load, the difference in slight over shoot on the inverter waveform is clearly visible on a DSO and accounts for this, I expect this with my build as I only have half the Power board ferrite beads, and this is likely a contributing factor on the Test inverter. Disclaimer: I'm not finding fault with the Controllers, they are doing exactly what I expect, or even better at very low currents through the sense transformer. I ignore any current readings below 1A or so on the LCD, there is a setting to make the LCD show 0.0 at Idle for any current offset, that setting does not change the accuracy of the sensor at higher currents in any way. You may get a different result in your inverter, many variables in build, layout, noise etc. But giving my results might ease any thoughts of a problem in the build, where clearly there is not. This is a powerful inverter and these very low currents are of no interest, well not to me at least, I'm not powering light bulbs. I'm super impressed with the DC voltage tracking and 3.0v volts to 60.0v is spot on. One thing I should point out in both my controllers (with R6-Inv divider value fixed), the voltages with one calibration are not really accurate enough for my code in my two inverters. It will work with one calibration, but if single calibration causes Vcap to read more than 500mv below Vbat, the controller may tell you that it's waiting for Vcap to charge fully. This can be changed in Setup, but really, Vcap and Vbat should display within 0.2v in a correctly built inverter with no bad connections or faults, both should be exactly the same voltage (with a DVM) before the Kilovac is closed and SPWM soft-start begins, again your inverter may have tolerances that allow one calibrate setting, and that is available in setup as well. I have a very specific reason for this - and it's glaringly apparent with a Peak DC input current meter on Kilovac enable, but it's also somewhat inverter dependent on the size of your capacitor bank value and wiring. Both inverters have been perfect so far, I've abused them, seriously so, and I cannot get either to glitch in any way, I haven't tried a 1kw RF linear next to them, but I'm sure user Dex will do a test and let us know  . Edited 2024-07-01 20:10 by KeepIS It's all too hard. Mike. |
||||
| -dex- Regular Member  Joined: 11/01/2024 Location: PolandPosts: 99 |
We will have to wait for tests with 1kW RF amplifier, but I will certainly let you know when I do it  Have you done any tests with connecting the GTI inverter to your off-grid inverter? Edited 2024-07-01 22:29 by -dex- |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1679 |
No I haven't, I have no need for it, but Murphy's friend may know as he built his own design based on some of wiseguys power stage circuitry, so he and wiseguy would be the ones to ask. I take it that your inverter is still running fine? It's all too hard. Mike. |
||||
| -dex- Regular Member Joined: 11/01/2024 Location: PolandPosts: 99 |
Yes, it works well, but it is not heavily loaded, so far no more than 3-4kW. I'm off-grid all the time, except for the heat pump that heats hot water. The heat pump is three-phase and it is the only energy receiver powered by the grid. I need to change it so that I can also heat domestic hot water using an inverter, for this purpose I need to install an additional heater in the tank and turn it on when other heavy appliances are not turned on. Considering that I like cold baths and winter swimming in general, it will be not so soon  |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1679 |
I have been reworking the corner of the shed that houses the inverter and solar regulators, so we have been on grid power for two days, late yesterday I finally had everything ready to switch back to Solar. I finished up and threw a switch to signal the ATS (automatic transfer switch) to switch to Inverter AC. Some good news and BAD news. Bad news first .... The inverter failed. The good news .... This time I was standing beside it, and knew what happened - but why now? This has happened once before, but when switching from Inverter to AC under a high load. The hint, as before, was the mains circuit breakers tripping in two locations, the bang I heard was the Kilovac dropping out, the over current LED was flashing, the other thud was the mains breaker tripping. I walk over to the House and find that my wife had the drier running on high, and the Hot water system was still switched over to top it up, so mains load was around 6kW when it switched from mains to the inverter. The inverter normally has no trouble with this. The first time it happened I had immediately suspected the ATS might have arched over - Mains AC and Inverter AC don't mix in this way. I ordered a different type of ATS, and as sh*t happens, I never got around to changing it over, but it was always in the back of my mind and was the next item on the to-do list. The Inverter took all of 15 minutes to get up and running, as I simply swapped in the new power board that I had recently built. A quick look and absolutely no sign of blown FETs, no smell, nothing, but three FETS were shorted - love these Kilovacs. So today, I had to find a way to incorporate this different type of ATS - and to pull the old one apart and see if my suspicions were supported. I should have ripped that old ATS apart long ago. Inside I find two relays, each rated at 60A, one switching Neutral and one switching Active - simply connected in parallel, the gaps are TINY, what could possibly go wrong above 5kw, it's rated for over 11kW.   Below The newer low cost ATS has a gap measured in inches - a rotating cam, huge contacts and dual solenoids. I'm using the 2 pole version of this similar unit for now.  There are dozens of the videos on the internet, some are wrong and would burn out a solenoid in no time, none show any use for the passive relay contacts apart from switching indicator lights, and the dam circuit supplied with them is wrong, and circuits on the internet are also wrong - WTF. None mentioned the warning about burning out solenoids if the AC was low, the manufacture does not hint why, but one look inside and it's bleeding obvious, I should have realized without opening it, but when your doing 100 things at once - sigh. So, I have it working with the addition of a single small 3A AC SSR "Solid state Relay" I used the same Mains/Solar priority switch with a tiny 240vac to 6v transformer and small housing I had used before, the only connections to the inverter are the AC enable output pin from the controller FET and a ground connection - brilliant! The circuit shows two relays, one on the Main power input and one on the solar/Generator input, this is wrong, there is only one relay for the two "Passive contact outputs" on each AC input. That relay is controlled by the Top input only, because of that, it allows the use of lower input passive contacts of that relay be used to stop the backup AC input powering the solenoid if the Top input also has AC mains. You cannot have AC on both inputs unless one solenoid is disconnected. The solenoids are driven by two wires from each AC input connected to the AR, AN and BR, BN terminals on the two green plugs. When a solenoid detects AC, it pulls the switch cam across and a micro switch disconnects the Solenoid to stop it burning out. You can see what would happen if both inputs had AC applied. It turns out though, that this is a really brilliant design and so easy to use once you realize the dam circuit is wrong. I can post the connection layout I used if anyone is interested. . Edited 2024-07-05 20:20 by KeepIS It's all too hard. Mike. |
||||
Murphy's friend Guru Joined: 04/10/2019 Location: AustraliaPosts: 648 |
Hmmmm, there is usually a price to pay if one makes things just a little too complicated (automatic?) I just use a manual changeover switch if I ever need to use street grid power. Not very difficult to ensure there is no load on when I switch back to inverter power. |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1679 |
That is true, but the these ATS devices are designed for this, the one I had for years was not up to the task on high loads, as others have reported when I searched for the old ATS, a number of cases reporting inverter failures from arching, the design was crap. Automatic is simple with these new ATS unit, there are no extra components needed. Mains stops, switch to inverter (if it's running) or back to Mains when it returns. My slight mod was to simply allow the priority of Solar over mains or mains over solar. The new ATS has a manual/Auto switch, so it can be manually switched if you want. As I indicated, it's not the load that is the problem, it was that old ATS arching to Mains AC when as it's connecting to Inverter AC - that can NEVER happen again, it's impossible in this new ATS . It's all too hard. Mike. |
||||
wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
This might be a reason to finally use a sync input ? If there are devices with reactive loads transferring, it might be more gentle than hard switching between sources say 180 degrees out of phase if they are running asynchronously ? I still havent quite got my head around transfer switches and delay times re switching etc I think I need to get one and do some reverse engineering to see exactly what makes them tick. I would like to have one but like MF I am a bit wary of auto transfer atm. Edited 2024-07-06 09:35 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 |
These can be used in manual mode, but also allow you to employ the solenoids to fast switch if you want. Or simply turn the knob and use it as a very powerful switch, with 2, 3 or 4 poles, 3 phase. They are really simple - I can post pictures of the inside showing everything if needed. I measured around 18ms "average" transfer time, but like any mechanical relay that varies, they are rated at less then < 50ms transfer under solenoid switching. Even with phase switching, I wonder how you factor in variable switching time, would you not need electronic transfer switching with a known constant transfer time. If people are using manual switching, are you only switching when the power mains has dropped to zero, or are you switching live? if live switching, the only difference is that the Electronic Transfer switch can usually switch faster. I manually switched the other day when that crap Transfer switch had a flash over, mains and inverter were live (AC up on both). It's not the ATS that's the problem, it's the design on that crap unit. . Edited 2024-07-06 10:26 by KeepIS It's all too hard. Mike. |
||||
phil99 Guru Joined: 11/02/2018 Location: AustraliaPosts: 2135 |
KeepIS.s old transfer switch might comply with the regulations for ELV (up to 50V RMS sinewave) but not for LV. For an automatic transfer switch simple changeover contacts are not permitted, at a minimum they need to be on separate armatures to ensure adequate isolation between the two inputs. Normally two separate contactors are used with both electrical and mechanical interlocks. The electrical interlock ensures the first is off before the second can be energised while the mechanical interlock ensures, in the event of welded contacts etc. the second can't close until the first armature is released. With regard to synchronizing the inverter with the mains, this is common on large 3 phase UPSs. As long as the mains is within about 0.2Hz of normal it is used as the reference for these inverters. It makes for seamless changeovers, especially if there are induction motors running. They can produce a large current spike if there is a sudden, large phase shift. |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1679 |
The manual is in Dutch and English, the manual states 240VAC, but that does not mean much with a compliance, it has a CE sticker but that means nothing. The new one is not a relay, it's physically impossible to have a changeover contact fault. EDIT: This is why I was amazed when I opened up the old one, no interlocks of any kind - absolute crap, it was raved about on line a few years back. Edited 2024-07-06 10:50 by KeepIS It's all too hard. Mike. |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1679 |
Just a clarification, when the Inverter failed due to the AC mains and the Inverter AC being connected together by the faulty ATS, the loads were all resistive - I don't switch when induction motors are running, AC failure will immediately drop all NVR switch gear. It's all too hard. Mike. |
||||
Revlac Guru Joined: 31/12/2016 Location: AustraliaPosts: 1024 |
Been a bit crook, still getting over this bug I got from shopping last time, anyway I have always preferred the mechanical (cam type) change over switch off in the middle position cant go wrong except for a lightning strike, I have some od 3 phase switch gear and its not possible to have the 2 big contactors on at the same time, it can only be one or the other because of a sliding bare that prevents it. Anyway was going to ask what your thoughts are about a common mode choke on the inverter DC power input? maybe 1 or 2 turns would be a big choke I would think, just wander if that maybe helpful or I'm talking rubbish and its not a good idea. Oh, Its great that your inverter is back up and running.  Cheers Aaron Off The Grid |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1679 |
Hi Aaron, a lot of that going around in certian areas. I did try that idea with a higher frequency choke but it made no difference to noise on the DSO, so that was promptly scrapped, and it needs to handle a few hundred amperes as well. The inverter was running about 20 minutes after the old ATS flashed it over with Mains AC, a little more then the time it took to swap the power board. No physical damage to the power board, just 3 shorted fets, no smoke, not even a hairline crack in the FET cases - AC and DC input current trip and a Kilovac to the rescue. FYI: The GAP on the first ATS.  GAP on the ATS I'm now using.  What about one of these solenoid switches  It's the weight of a small vehicle battery.   . Edited 2024-07-07 11:30 by KeepIS It's all too hard. Mike. |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1679 |
In a previous post I indicated that I had found 3 FETS shorted on the power board, I decided to use 3 of the undamaged? FETS from the previous episode with that rogue transfer switch. Now this is frowned upon and rightly so, however I've got nothing to loose in seeing how far I can push the Test inverter, and if any of the FETS are/were compromised. Before I'm flamed - YES - I know - just because I might get away with it this time, does not mean I could the next time. A quick swap of 3 FETS and powered up in test mode, no caps, 1A current limited supply at 15V - bad waveform but some AC out, power supply current trip flicking but no shorts, then I remembered that I had not measured the Gate 1.2ohm resistors on the blown FETS, sure enough one was open - same thing as last time but then two resistors were open. Replaced one resistor and the power board is running in the test inverter, the floating gate of a FET randomly conducting in the inverter power stage is not nice. I'll do some high current testing when I get time. One of the great additions in this Nano controller is the AC Mains up control pin on J9. That control line is driven via a FET and pulls to ground when AC output is up. I use this line to control the ATS - the instant the Inverter is switched off, the AC ready line is released, or if the AC output drops below the AC low voltage point setting. This totally simplifies the Automatic transfer to Mains in the event of the Inverter stopping for any reason, low battery, current trip, switched OFF ETC, and is almost seamless except for one flick of the workshops 9 Fluro-LED batten lights. . Edited 2024-07-08 08:33 by KeepIS It's all too hard. Mike. |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
A few questions, were the 3 FETs on the high side or the low side or both ? I have no trouble with the decision to keeping/recycling the working FETs, I usually isolate the gate of what appears to be a working FET, momentarily charge up the gate with 12V (then remove the 12V) and then check D-S is less than ~ 0.05 OHM (for a typical power FET) and stays that way for a while. If the internal gate junction is not damaged (in which case the D-S will very quickly become open circuit again) the FET is likely to be good. I have been replacing just the faulty FET/FETs in parallel power configuration for more years than I care to count, we also kept meticulous records of the faults and cures in a comprehensive database compiled for over 10 years of service so if we had an ornery problematic charger, it was then decommissioned and replaced. If there was a follow up FET failure after a repair, we would know, as the whole history of that charger was presented when updating the database. I would recommend always changing the Gate drive resistor of a dead FET even if it appears to be about the right value with a DVM. I would go further and suggest that the buffer transistors may have been damaged also, do not rely on diode/resistor checks of the transistors, I have seen cases where measurements seemed about right but the devices were actually faulty. Do use a CRO if available to check the Gate drive waveforms at the actual FETs G-S after any repair, looking for rise and fall times of 50-100 nSec. Testing with a current limited lower voltage test (10-16V) on the FET Power Bridge is definitely a good plan after a repair first, then increasing the voltage to ~50V to check overall performance if the LV test looks ok, before reconnecting to the Battery sources. If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1679 |
I always check the waveforms for each stage from the Opto to the FET gate and A, B outputs. If both low voltage and full voltage to 60V pass, I then add the Test cap banks and run some load testing before fitting the Running set of Caps boards, so easy to do with this design. Good information on the quick test for the FET Gate, also nice to hear about the logged history of faults, then you know you can trust the results. . It's all too hard. Mike. |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
And my original question - were the 3 (dead) FETs on the high side or the low side or both ? I am curious exactly what damage resulted from the crappy transfer switch event (to try to develop a theory as to why). The comments re FET testing & replacement were mainly for anyone who is not really familiar with FET catastrophes. The comments also relate to subtle FET damage that is not consisting of multiple exploded cases and molten FET leg & guts remnants - best to replace them all for these disasters (including driver buffer transistors). Edited 2024-07-08 14:10 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 |
I wrote the last post just before I had to head out for surgery, and remembered you had asked that question but alas too late to answer. Back now, so one on the Low side, which also took out the gate resistor, I'm assuming it was the first, and a high and low FET on the other half, Q12, Q5, Q13. Inverter was running with 40,000uf, and as indicated, the inverter was not disconnecting from a few kw of mainly resistive load, it was connecting to the load via a POS claimed 11kW rated changeover switch. It's all too hard. Mike. |
||||
| Page 10 of 23 |
|||||
 Print this page |
