 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 : Hopefully? Another 48vdc-240vac Toriod Inverter build.
  Page 12 of 26   |
|||||
| Author | Message | ||||
| analog8484 Senior Member  Joined: 11/11/2021 Location: United StatesPosts: 108 |
Thanks for the explanation. This reminds me of a question that's been in the back of my mind for a while. The question is about the relationship between the DC bus capacitance and the choke value. It seems to me a key reason for the choke is to improve DC bus voltage stability. So, assuming a perfectly stiff DC bus using a huge capacitor bank with appropriate battery wiring then the choke can be smaller? |
||||
| nickskethisniks Guru  Joined: 17/10/2017 Location: BelgiumPosts: 458 |
Wiseguy, well explained! Analog8484, Not completely, you want a stiff DC bus to avoid side effects. One of the reasons you want a stiff DC bus is to compensate (maybe not right term) your wiring/circuit inductance. On first hand the inductance of wiring and circuit looks small, but the faster we are switching and the higher our voltage difference, the higher our di/dt, resulting in higher voltage swings on the DC bus (and between Udrain and Usource of the mosfets), possible oscillations. If you don't compensate this, you can get huge voltage swings and destroy your mosfets, or having trouble in the output regulation and gate drive. A big capacitor bank helps to prolong the life of your capacitor bank, as hopefully the current ripple and heat dissipation is evenly spread over the capacitors. Keep in mind the stiffer everything gets the more complicated things can get, I think that this is something KeepIS is meaning with, we build bigger and bigger, better? (faster and faster) gate drivers. Rise times maybe get dangerously fast etc,.. and invoking actually more or other problems we never had before. But like a colleague is saying, there is always a reason things fail.  But we don't always want a stiffer bus, I've seen some examples in dc/dc converters (small < 12VDC and even 3KV) where they put a series resistor to make the input impedance worse, because of stability issues. In theory there is no relationship between your choke value and DC bus capacity, Wiseguy did describe how you can calculate the choke value very well. You are right there is a positive side effect, the reduced peak current. But you can't really say the bigger your DC bus capacitors the lower your choke value. You can dive more in to LC-filter design if you want to learn more about the subject. Actually, if your bus has a very low esl,esr, etc... the higher your potential current can be. And the more you need to be careful, because if you reduce your choke value then you get faster current slopes and hit your saturation point much quicker. Wiseguy, is it possible the +1000A pulses you described are lower due to the leakage inductance of our transformers? (I know you said you are ignoring some factors, but I felt I needed to mention this) On the other hand, our transformers will have a certain capacity as well what makes things worse. Even our choke, but as you said don't make things over complicated.  Sorry, I think I wanted to say that my own experience learns me the more I know something about it, the more I realize there is so much more to learn about this topic. There is no end in this, I don't think we need to understand everything but on the same time it makes me crazy not knowing and motivates me to keep learning. I hope to make more time in the near future to share my findings and testings. I even quit my management function and switched back to my old job as semi repairer power electronics and build test environments again, because I want to have fun and learn again. Edited 2023-05-06 05:28 by nickskethisniks |
||||
KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1679 |
Yes, I should have worded that a bit differently. I was alluring to all the voltage and current parameters that should be taken into consideration for the inverter build. Like low cutoff timing etc, and to keep as a reference for working parameters once the Inverter is set up and running correctly. Then if something changes, or as a general maintenance check, I have a known logged set of working references to test, verify and fault find with. Mike fully acknowledge your non-guarantee, but it's the extra insight and the interest it gives some of us to look further into that area of inverter operation, I really appreciate, or put simply .. It makes us think. . It's all too hard. Mike. |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1679 |
Absolutely how I feel  Also thanks for that informative last post and extra insight. . Edited 2023-05-06 08:16 by KeepIS It's all too hard. Mike. |
||||
wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
Thanks for your post Paul, It's nice to see some confirmation of my theoretical results. I will also need to rework/tweak my chokes and I suspect the "grunt" issue I had will vanish. I also did not mention that the chokes storage ability is dependent on the cross sectional area of the core, core volume, gap length and the core material. Simple hey ? lol  If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
Hi Nicks, thanks for the comments and answering about the bus capacitance/choke value question. The leakage inductance can be considered as another small choke increasing the value of our series choke by an amount that is equal to the leakage inductance. Some manufacturers actually wind the toroids by design to have a highish value of leakage inductance and then don't use a series choke at all. So the answer to your question is yes leakage inductance and series resistance of the winding will all reduce the very high current values I posted - that was really more to emphasise a ball-park scary number. I also apologise for the 0.014 milliohms in my recent post, of course it is either 0.014 ohms or 14 milliohms....can't be both. If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
Sorry for polluting your thread Mike. I have reposted this section and removed the RMS values that were shown after the 52V Peak current values, as they were wrong. The RMS value should have been the same as for 45V, I took the new higher peak current value and multiplied it by 0.7071 to get RMS - you can't do that with square waves only sines ! For a 1kW output: DC battery input current at 45V is 25A, at 52V is 21A Primary AC input voltage Peak is 41V, its AC RMS voltage is 29V FETs Peak current with 45V input is 54A or 38A RMS At 52V DC in, the FETs must still supply the same energy but they are now operating at a lower duty cycle, else with the higher input voltage for the same on period the energy would be greater so with 52V in the FETs Peak current is 62A For a 2.5kW output: DC battery input current at 45V is 62A, at 52V is 53A Primary AC input voltage Peak is 41V, its AC RMS voltage is 29V FETs Peak current with 45V input is 135A or 96A RMS FETs Peak current with 52V input is 156A For a 5kW output: DC battery input current at 45V is 123A, at 52V is 192A Primary AC input voltage Peak is 41V, its AC RMS voltage is 29V FETs Peak current with 45V input is 271A or 192A RMS FETs Peak current with 52V input is 313A Edited 2023-05-06 09:39 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 |
This is absolutely not pollution Mike, any insight, clarification or thoughts or any aspect relating to making an inverter that will last is appreciated and welcome. . Mike, in your EG8010 Rev2 controller, you have 3 pins marked SS2, with D13, Q4 and Q5 + an IFB connection, what is wired to, or jumpered on, those connections? Thanks. EDIT: I just realized my description of the battery drop was not relayed to you correctly, and your interpenetration of same as it's slightly less. The nominal battery voltage when Solar is supplying everything or under light loads is around 55V, but once you start drawing a good load, and running on battery, it sits at 53.5V. The battery basically sits within a few mV of that all night under load. The 55V is basically surface charge, but that term may be incorrect for LiFePRO4? So in realty the battery voltages drops from around 53.5v to 47.8v, so a 5.7v drop at over 520 amps between the battery terminals and the inverter PCB +/- connections. So, all switches, fuses, connectors, contactor, cables and battery internal R, for both Positive and Negative paths combined. . Edited 2023-05-06 10:22 by KeepIS It's all too hard. Mike. |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1679 |
I wanted a small DSO for quick LF, audio and digital work. These were pretty cheap, rated to 500KHz, but surprised it's only very slightly down at 1Mhz. It came with a good 100mHz CRO probe and some other features I didn't even look at. Now, I know there have been tons of testers around for years, but I never bothered as I just used a multi meter and they used to be a bit pricey for limited use. So apart from the usual inductance, capacitance, resistance, zeners to around 30V, some IR sensor decode and temperature sensors, it comes with a nice clear indication of some unknown devices. The DSO screen is good with plenty of options, great for monitoring the Inverter AC, and it has a long battery life. But this came in handy today whilst going through a box of unknown devices, so dam fast and easy.  . It's all too hard. Mike. |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
I will look at those connections for you later tonight- laying floors today  If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1679 |
Thanks, that would be great. I'm slowly getting the control board build, then I can play "kill the inverter" all over again, and see if I loose this time  BTW I finished replacing a backyard deck a few weeks ago, that was bad enough, but flooring is a royal pain. Have fun  . It's all too hard. Mike. |
||||
| pd-- Senior Member Joined: 11/12/2020 Location: AustraliaPosts: 122 |
It must be hands & knees week i am tiling a floor |
||||
| analog8484 Senior Member Joined: 11/11/2021 Location: United StatesPosts: 108 |
Thanks for the explanation. I have also wondered about many of the factors you mentioned and that's why it wasn't clear to me if a stiffer DC bus would allow smaller choke. The lower ESR is a particular potential issue that I suspected. And a related topic is the much lower ESR in the change of battery chemistry from Lead-Acid to Li-ion. My generally feeling is that lower DC bus ESR reduces the safety margin for error (design or manufacturing). In any case, it seems DC bus stiffness can definitely have impact (positive or negative) on the choke value and overall LC filter design even though theoretically they should be independent. I suspect such things contribute to the trial and error nature of choke/filter design. A more comprehensive real-world model of the entire power chain for inverter design seems to be missing or I haven't been able to find one. |
||||
| analog8484 Senior Member Joined: 11/11/2021 Location: United StatesPosts: 108 |
I also read that somewhere but it doesn't seem to be common. I wonder what is the downside of the approach vs separate chokes. |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
Firstly re the edit, this is a family friendly forum and I'm not into interpenetration which sounds suggestive and maybe even painful..... Now more seriously, the EG8010 is I consider very inadequate with regard to data and specifications for the various pins control and function pins. It also has provision for connection of an LCD which displays volts amps temperature etc. This is the data sheet I worked from - if anyone has a better one around please upload or link it. EG8010_datasheet_en ser.pdf The Amps input (IFB pin 14) for the 8010 is derived from a resistive shunt which is shown as 0.1 Ohms for the 400V switched choke output (no transformer) or 0.01 Ohms for the Transformer output which according to their schematic can be fed from 12, 24 or 36VDC. But for a given load of the 12/24/36V supply would have 3 totally different currents showing for the same load so unsure what the display would actually read. There also appears to be no way of calibrating the current feedback. The IFB pin is simply described as 0.5V typical, and that if exceeded it will shut-down and restart 5 times before needing a hard reset start if normal operation was not resumed in the 5 times. Where the shunt is placed at the bottom of an output FET bridge of the transformer output stage, I believe it just represents a reflected current proportional to the AC load. So to your original question I allowed for some investigation and experimentation and provided a link to the current sense stage when the link is in the 2,3 position. I thought maybe the display could be useful but was unsure how the current was to be interpreted and displayed. When the link is in the 1,2 position it just shuts down the output when the 2 transistor over current latch has been triggered and needs a reset of the latch via a push button. I suggest link 1,2 for normal no display use. I must warn you though that I have hardly played with this PCB version at all except to make it start and create a regulated output. In hindsight the over current trip and the following push button over current restart should be verified to ensure that it actually does soft start and if not a link to the start pin circuitry feeding pin6 might be needed to ensure it does a soft start. Let me know - I cant remember, I think it does do the soft restart after a restart / overload event. Times up back to the flooring Edited 2023-05-07 12:02 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 |
Mike, all good and understood, I have been going through 8010 document again. I tested the control board and was greeted by a flashing low voltage code, which I thought was great, it lives. I fed it a 50Hz sinewave from my generator and checked I had 100HZ half wave signal out of the Pre-Rectifier. I could not make the Error code go away though. I have a few of the small LDC display units for the 8010 boards, these fit to a header on the 8010 boards. I decided to see if one would work here as it might save me some time with fault finding controller status. Fortunately I wired in the LCD header when I built the board, and I'm sooooo glad I did. I rewired the pins on the LCD cable to suite your layout. Result was a neat display showing 50Hz, temperature over range ,fitted a 10k and got temperature to 24deg, all appeared good. Fed in my 50Hz and it showed the expected AC voltage on the display, but I could not stop the under voltage Code. I checked the PWM outputs and all were working correctly, and the inhibited state logic was correct. Connected a PSU to the Power unit, a small 12 to 240v toriod, a choke and a some small value cap on the power board and fired it up (resonated Toriod first). It worked first go, 240vac output and soft start working correctly, tried it with over temp, and dropping 12v DC power to Controller, it shuts down correctly, soft starts when 12V power returns, or if testing over temp, when temperature is back to normal. AC wave form was a nice sinewave but a slightly rounded due to almost no CAPS. So now I can set it up to Latch and not power up. So unbelievably, I built the Power board and the Controller board over two days and the whole dam thing appears to have worked first go The only problem is I need more capacitance, both my supplies go bit nuts with incorrect voltage shown supplying the Power board, and I cannot adjust the AC output, however the Controller and LCD correctly show me changing the expected output based on the VFB, so I think it's down to too little capacitance, the small cap I was using for the test was getting warm. Expected low current across both board though. . It's all too hard. Mike. |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1679 |
I fitted each test cap board with around 3000uF, for a total of 6000uF, that 4000uf less that just ONE of the eight big caps I'm going to use. Much better behavior from the Power supply now. I can now adjust the AC Voltage, BUT the little LCD stays at 219v now, it climbs up to that at power on and stays there? Maybe the 8010 Firmware is doing more with AC input then we think and 100Hz is upsetting the voltage calculated for the LCD display? The VFB pot is nice and smooth and regulation under load is fine, and as I said, it adjusts 230vac out fine. I only ran a small load as I'm limited to PSU 5A current trip at the moment. Left it running for an hour, all cold and happy so far. About to connect the Current Xformer.    Unknown transformer, did not test turns ratio, nominal 13v to 240v toriod on label. No AC out filter, one choke and 6000uF caps.  . It's all too hard. Mike. |
||||
Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
The EG8010 is VFB expecting 3VDC as per this table in the datasheet. Just use a 12v or similar transformer, rectifier and voltage divider and you're in business.  There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1679 |
Thanks, I have already done that to confirm the 8010 and basic circuit operation, however this circuit and its operation are completely different. I have it working but the LCD unit changes what it wants to tell me about VFB, so this LCD issue is of academic interest only, as it in no way affects the Inverter. BTW I have the full EG8010 Datasheet. But that may be of interest to someone who has not seen it. . It's all too hard. Mike. |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
Mike, you have done well the waveform looks pretty good for a first turn on. As madness pointed out the VFB is typically ~3V when operating normally. The small PCB voltage transformer and full wave active rectifier should be delivering ~ a train of 3V half sinusoids, resembling a full wave rectifier waveform with no filters, to pin 13 of the EG8010. I'm not sure whether you have an issue or not at this stage, but please get back to me before Sunday next week if you do, as my emails etc will be few and far apart for ~ 5 months, and I wont be able help by by bench testing anything until we get back. I think the AC output voltage may be a bit low at 220V to capture and regulate the output. Turning up the supply to ~ 55/60V shouldn't do any harm to anything, to prove a point. At the moment you and I are pioneers and have the only alternating unipolar drive using inverse optocoupler drive from an EG8010, I suggest in the world lol. I will add that you have probably advanced a bit more with yours than I have with mine to date. I still want to see how it performs in a serious application yet but it will be a while... BTW have you got a link where you bought the little CRO from ? If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| Page 12 of 26 |
|||||
 Print this page |
