 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 : Controlling a GTI into the ozinverter
  Page 3 of 6   |
|||||
| Author | Message | ||||
oztules Guru  Joined: 26/07/2007 Location: AustraliaPosts: 1686 |
yes thats the line to alter.. after setting the pots.. one for course, and one for fine. It needs long instead of int for the time counters or they o/flow... we'll get to that too. Just a simple voltage divider, feeding the nano with control voltage not exceeding 5v ( zener on input).... simple linear reg to feed the nano and opto drive. The rest is in the software....simple too. Then the output from the nano goes into a A3120 opto.. that then drives the remote fets. It easily is able to drive 8 big fets with high capacitance inputs cleanly... as we are only going at low frequency.. The scope has no vertical component in the sweep, as it is so fast in rise and fall times. So it can be used for GTI or normal voltage control. Embarrassingly, I made a few changes, and saved it, and now it won't print to the second line for some reason ... now will have to find out why... makes no sense to me, why I cannot for love nor money print to the second line anymore  .......oztules Village idiot...or... just another hack out of his depth |
||||
| oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686 |
I solved the problem ......by running away.. I could not find why the LCD would not under any circumstance print to the second line... so I ditched the thing and started from an earlier one I had posted.... no idea where the gremlin was. Line for line it seemed the same, but ran totally different. Mad, I decided to use your idea of serial communication.... but cant figure out the web of libraries and which to use with what displays... is it as bad as it appears.... or are there generic ones...... ......oztules Village idiot...or... just another hack out of his depth |
||||
Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
Hi Oz, Yes they can be tricky if you don't have the right library, look here covers most possible types. There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
| oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686 |
I'd like to tell you how wonderful my programming skills are.... but alas, a few bugs turned up... kept reverting to bulk under certain circumstances. Think it's fixed now.... but will wait a few more days before I get even a little bit cocky. Surprising how such a little board can cause so much consternation.... to people like me who don't know what they are doing. I'd like to think I was learning a bit more all the time.... but it doesn't show up like that..... just seem to dig a deeper hole sometimes. ..........oztules Village idiot...or... just another hack out of his depth |
||||
| Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
I would prevent it going back to bulk/absorb for a period of say 10 hours or if the battery voltage dropped below say 52V for more than a set period. I have been meaning to order some FETs to build one of these, the one you used is $46 from RS. I will need 2, one for each channel of the GTI. There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
| oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686 |
Wow.... and I mean Wow..... thats a lot of dollars for a fet. Even better would be these http://www.ebay.com.au/itm/2pcs-IXFH60N50P3-IXYS-MOSFET-500V-60A-0-1Ohm-PolarP3-Power-MOSFET-/262543513005?hash=item3d20 d015ad:g:XYgAAOSw0fhXlWSx or There are plenty of alternatives for peanuts that can be doubled up. I rewrote it more sensibly, and it seems to be stable as of now... but a full day of sun and draw will make for a better understanding.... testing on the computer is not the same, and has led me up blind alleys before today. Switching is less than 500hz I think, so switching losses are puny, particularly with the latest iteration using the opto drive. Wait a day or so and see if a better one turns up out of the testing... all looks very good at the moment. Most trouble was with he 48v version, not the 500v version. .........oztules Village idiot...or... just another hack out of his depth |
||||
| Mulver Senior Member  Joined: 27/02/2017 Location: AustraliaPosts: 160 |
For the voltage stable 8010 oz inverters would it be possible to digitally control the inverter output voltage? for example set a base voltage 230v then use an ardunio controlled digital pot to then manipulate the voltage up and down on the inverter based on battery voltage. 230v Normal overnight voltage / give me everything you got. 231v Seeing battery voltage rise/increase Pulse width 232v Bulk voltage achieved / hold pulse width 233v Decrease pulse width 234v starting absorb / increase pulse width 235v absorb achieved / hold pulse width 236v decrease pulse width. And so on??? it would need to be based on average/hysteresis time. This would mean remote grid ties can function with no communication but the mains voltage. It could be continued to float and then to show excessive power after batteries full and to turn other things on hot water/heaters/ac?? |
||||
| oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686 |
Yes you have complete control of the inverter output voltage if you over ride the settings... But remember you will lose the ability to compensate for voltage fluctuation via loads or GTI inputs, so it will have to be a pretty good program in the arduino, running very tight timing to keep the thing in control.... your reference to hysterisis and averaging would be necessary...... however.... If you use the arduino for offset compensation, then it will be very easy, rather than taking control... ie just move the base line that the 8010 is using to control all the transients. This could be as simple as a series resistor that you increase /decrease the off set over, and that will fool the 8010 into a false voltage range, without impairing the feedback system... ie add more or less offset to the normal control signal. Sounds like you have a cunning idea... clockman will like this if you can pull it off well. ...........oztules Village idiot...or... just another hack out of his depth |
||||
| Mulver Senior Member Joined: 27/02/2017 Location: AustraliaPosts: 160 |
In regards to electronics I'm absolute beginner level, but offsetting sounds relatively simple!! Some further simpler thoughts on required signals ..... -3v turn off non essential equipment -2v more grid tie -1v less external load -0v hold (base set voltage) +1v more external load +2v less grid tie +3v fire up the welder |
||||
| oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686 |
Your definition will be fairly low as you will have to translate from 340v peak down to 5v max for the nano to read. The good news is that there is a significant range that you can operate the AC voltage over befor anyone really notices big time. If you keep to less than +-10%, you will be ok, and that gives you a range of 40 odd volts really... thats how forgiving the power systems really are... so you can realistically get up to 260v before getting hairy, and down in the 220v range. That kind of range will give you a better chance of pulling it off consistently. I find the 8010 can keep in a 2v range at almost any normal power level.. ( <4-5kw or so ) truly amazing how well it regulates. Temperature will also allow for drift of a volt or two.... so it is not as tight as you imagine... although, once up to temp it seems very very stable. Mad or clockman may be able to add more empirical data to that. ........oztules Village idiot...or... just another hack out of his depth |
||||
| Mulver Senior Member Joined: 27/02/2017 Location: AustraliaPosts: 160 |
The other problem is the reaction time of your controller i imagine is quite fast. And swinging the grid voltage all over the place to control pulse width is probably not a great idea. So I'd think it being a more overal signal saying hey guys give me more or give me less. The problem then is the rapid pace a grid tie changes output with clouds then blue sky causeing rapid back feeding! So maybe a current sensor on the grid tie output to your nano so you can offset the pulse width instantaneously when clouds cover or sun comes out. Example. Your pwm at 50% seeing 1000w. The grid tie then goes up to 2000w. So you offset your pwm to 25% putting it back at 1000w Edit. The above example is over simplified on further thought. But I think the theory is sound  |
||||
| oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686 |
We don't need to be particularly fast, the readings and corrections are going at 7 times per second... thats way more than we need. The increments are fairly corse, and they seem smooth to the GTI... We don't care what the GTI sees, we only need to care what the battery sees, and that is relatively speaking... time stable. It has a fair inertia, but with a few kw flowing in and out can be volatile... but still the inertia helps immeasurably to keep things calm. So for control, we only need keep the battery voltage below a certain level ( 59v and 57 volts) depending on the stage we are in. Now when we try to do it with AC volts... it is still the battery that we measure, and whatever error signal we need, gets translated to the GTI... exactly the same as the battery next to the GTI would. When we want to control other things by changing the AC voltage ... and we do it via offset.... we need only worry about the fixed offset..... because the main inverter controls the AC voltage, the GTI just follows blindly.... It only knows how to push back against the mains. If the mains goes up, it has to try to go up higher just to keep pushing back into it. The only time the GTI affects the voltage really is when it generates excess power, and the inverter can't use it... it will drive the battery voltage up, and the AC with it, as the energy has to be translated into watts somewhere... if the current can't go up, the voltage must.... it will try to keep a higher forcing emf than the mini grid is... or it cant feedback..... we are just choking the source power from it. Well thats how I see it working. .........oztules Village idiot...or... just another hack out of his depth |
||||
| Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
Oz would it be more effective to regulate the GTI via the inverter by increasing the AC output voltage as the battery volts get to the set point? Then it would not matter how far away the GTI is from the inverter. Just ordered some SPW20N60S5 MOSFETs for $1 each that I hope will be right for the job on the high voltage DC. There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
yahoo2 Guru Joined: 05/04/2011 Location: AustraliaPosts: 1166 |
Or we could have the GTI's controlled to a fixed voltage. That would get around the funny feedback loop I can see in my head. There would be issues with wire resistance limiting the GTI output at high loads but I guess we could compensate for that back at the OZinverter. I'm confused, no wait... maybe I'm not... |
||||
| oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686 |
The figures look good Mad. Yahoo, we can't control the GTI voltage output persay... it will go as high as it needs to overcome any resistance to achieve getting rid of it's power somewhere to the best of it's mppt curve.... so the voltage makes no difference to it unless it goes so high it needs to shut down because of regulations. We can see that the controller we might put in the remote GTI box will have issues with line loss, but as mulver mentioned, we could measure the current and calculate the needed offset to compensate at the GTI.... we do have a micro in there I guess. There will be unseen problems with the AC control system, but I think we can still ,make something useful out of it. The DC control system works perfectly now.... thats a start. .......oztules Village idiot...or... just another hack out of his depth |
||||
| yahoo2 Guru Joined: 05/04/2011 Location: AustraliaPosts: 1166 |
That is not quite what I meant, I am saying control the solar panel DC input so that the GTI has a maximum AC output voltage that is roughly fixed say at 236V. Then vary the offgrid inverters voltage up or down a couple of volts to create the required voltage margin between the two to manipulate the path of the current output. There might be a technical reason it cant be done but the theory works. This is how I set up managed export grid tie systems on long rural lines, once the mains voltage rises over a set voltage the GTI's idle until a load is switched on that lowers the voltage and they roll back on. I am doing it to avoid the overvoltage shutdown where 4 or 5 inverters on a spur line keep crashing for most of the afternoon. it would mean the output control at the GTI is simpler coding and the output control at the off grid inverter is managed or manipulated a little more. I'm confused, no wait... maybe I'm not... |
||||
| Mulver Senior Member Joined: 27/02/2017 Location: AustraliaPosts: 160 |
Now that's an interesting idea !! |
||||
| oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686 |
No Yahoo, thats a simpler system, and one I didn't think of. The units I have built will easily do that control.... provided the hf units can accept the pulsed inputs like the transformer types can.... and I think they can. Solves a lot of problems easily now I look at it like that. ........oztules Village idiot...or... just another hack out of his depth |
||||
Clockmanfr Guru Joined: 23/10/2015 Location: FrancePosts: 429 |
Good morning Mulver, Firstly, ….. the PJ boards AC output approximately follow the DC battery voltage rise. The 8010 board only follows the DC battery voltage when the the Batteries are full, ie, the back charging has absolutely nowhere to go. However, on a few murky days, my 3 GTI’s. (soon another 2 ), are all putting there 20% ish, of rated power in to the batteries. The sun comes out briefly, the PJ boards AC rises as it has a back load, tripping out a GTI, but the batteries are wanting that backcharge. The batteries are NOT charged up.! ie, the PJ board AC is too sensitive and up and down like a …... The 8010 boards in this situation hold steady and allow all that sunshine created power to backcharge into the batteries, until they are full and I mean FULL and over cooking is not good for batteries. That’s if the sun keeps shining. With the above considerations…… ……… AC Load shedding, I use Hugh Piggotts method, http://scoraigwind.co.uk/tristar-follower-to-control-ac-heaters/ that runs from my DC Tristar load diversion controller. Works well for hot water, but over the past 8 years I have seen faults, and things can get mighty hot when you need to be dumping. ……….DC Load shedding, done with my 3 DC wind turbines, using 3off PWM Tristar diversion controllers, problem is, that sometimes these operate and try to dump the GTI’s output. Conclusion…… I do like what Oztules is trying to achieve on the DC side of the GTI, as I start to realise that my UNIVERSAL BOX OZINVERTER GTI CONTROL for each GTI, is not going to work well on the AC side. So at present its back to the DC battery state, either a MASTER control, as Oztules has created with charging algorithms, and SLAVES operating from a data cable or down the AC cable (please pretty please!), from the MASTER that shuts off the AC side on other GTI’s on the OzInverter.s Mini Grid. My philosophy, keep it as simple as possible, ROBUST and solid, and as cost effective as possible. Pics ... Fortunately I have the internal reset codes for these GTI's, so I have set these to trip out at the optimal 241vac, 242vac and 243vac, for testing the AC switching in real situations. You can see which of my used/second hand, cheap as chips GTI's gets viewed by the Public.......    ...... Everything is possible, just give me time. 3 HughP's 3.7m Wind T's (14 years). 5kW PV on 3 Trackers, (10 yrs). 21kW PV AC coupled SH GTI's. OzInverter created Grid. 1300ah 48v. |
||||
| Mulver Senior Member Joined: 27/02/2017 Location: AustraliaPosts: 160 |
Hi Clockman! Both pictures look pretty good but the top one takes the prize! I think there is possibility for a grid wide solution to your problem. If you use the 8010 inverter. Yahoo What you think of this? (236v just the number you used )So if we have a nano controlling the ozinverter output voltage (as per oztules comment) dependent on the actual battery voltage. Basically its own charge controller. Below 236v will allow back charging the lower the voltage the faster to a bottom voltage set point. We then have a nano in every grid tie moderating the grid tie voltage to a maximum of 236v at which it will start to throttle back pulse width to the DC not allowing it to go over voltage. This means all gird ties are independent yet providing for the same goal! Id expect the Grid tie with the most power would be the first to throttle back while others maybe in shade will be trying to squeeze every last drop? Would it then be possible to have "Loads" controlled with the same method as the grid ties but in reverse? A hot water service is an easy and useful one, a Nano the same setup as the grid tie nano watching the voltage but with the opposite maths at 236v and above it starts winding up pulse width to a solid state relay and below 236v it starts throttling back until it turns off. Do you think we will end up in a feedback loop in this situation? The goal being to turn other loads on when there is power available without grid wide coms! |
||||
| Page 3 of 6 |
|||||
 Print this page |
