 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 : Microcontroller and PC projects : maximite solar MPPT
| Author | Message | ||||
crez Senior Member  Joined: 24/10/2012 Location: AustraliaPosts: 152 |
If anyone is interested in building a MPPT solar charger around a maximite, I have adapted Tim Nolan's MPPT hardware and firmware to work on a Duinomite. It is running well on a 24v 180 watt panel but the plan is for it to handle 40 amps. I will post the circuit and program if anyone is interested. David |
||||
belsean21 Newbie  Joined: 06/07/2013 Location: AustraliaPosts: 12 |
Please do Sean. |
||||
Gizmo Admin Group  Joined: 05/06/2004 Location: AustraliaPosts: 5078 |
Please do too. That's a project I would be very interested in. Glenn The best time to plant a tree was twenty years ago, the second best time is right now. JAQ |
||||
James_From_Canb Senior Member Joined: 19/06/2011 Location: AustraliaPosts: 265 |
Will it work with other members of the Maximite family? James My mind is aglow with whirling, transient nodes of thought careening through a cosmic vapor of invention. Hedley Lamarr, Blazing Saddles (1974) |
||||
| crez Senior Member Joined: 24/10/2012 Location: AustraliaPosts: 152 |
James et al, Mine is running on a Duinomite at the moment, but should be fine on any maximite variant. I am spending an hour or two tidying up the software to make it less confusing. I also need to produce a readable schematic to go with it. I should have something posted in a few days. Part of my reason for posting this on the forum is to make me a bit more motivated, disciplined and professional.  It's working so far!
David |
||||
| crez Senior Member Joined: 24/10/2012 Location: AustraliaPosts: 152 |
2013-08-27_122723_MPPT24B.zip Hi Folks, Above is the zip file which contains the circuit and .bas file for the maximum power point tracking charger as is currently running on the duinomite with the latest 4.4 version firmware. The software works but would be nice with some logging or more graphing or better display. I will add some comments soon when time allows or if questions arise. David |
||||
| crez Senior Member Joined: 24/10/2012 Location: AustraliaPosts: 152 |
A few comments and corrections: Q6 and Q7 keep the duinomite supply low until the solar volts reach around 25v, at which point positive feedback from R12 makes Q6 snap on. They also keep the duinomite input below 26v. Without this circuit there are two problems. First the solar volts can go above 40v, too high for the duinomite supply. Second, the gradual increase in solar volts at dawn prevents the duinomite booting reliably. U3 amplifies the voltage across the 1 milliohm resistor by 60. This gives a voltage of 2.4v at an input current of 40A. Q9 and Q8 turn off when the pulses from pin4 of U1 stop. This prevents the battery feeding back to the solar panels at night, but also shuts down the supply to the duinomite until dawn. If I was totally confident pin4 of U1 will not get stuck in the high state, I would drive the gates of Q1 and Q2 from U1 pin4. As it stands, Q1 and Q2 are being used as diodes. This arrangement would not allow the bootstrap output of U1 to work on start-up, so Q5 is added with a series resistance to overcome this problem. The limited current in Q5 allows E2 to charge, allowing U1 to generate drive pulses for Q3 and Q4. If pin4 of U1 does get stuck in the high state due to a software glitch, L1 will light but nothing will fry. The efficiency lost by using Q1 and Q2 as diodes is not that significant on a 24v system. Q5 really need only be a low current device. At 40 amps, I1 needs to be a serious thing. I have used three stacked iron powder cores, the largest that Altronics sell, with eighteen turns of eight parallel 1.15mm dia wires. If it passes the test at 40A I will put some more details up. The software displays the input current and voltage and the battery voltage. It also plots the pwm duty cycle on the horzontal axis and power on the vertical. This allows you to watch the tracking process as it jitters either side of the best operating point. I have tested this on a 180w panel, with no sign of warmth. I will test it shortly at 40 amps and see how it holds up. ************** Errata: ************** 1. There should be a 470k resistor across C2 to discharge it when the pwm pulses stop. Actually, there is sufficient capacitance in the gates of the fets, so C2 can be left out. Don't leave out the 470k resistor though. 2. There should be a snubber network across I1. I am using 0.1uF 100v in series with 560 ohm, 0.5 watt. This reduces the level of radio frequency interference and also reduces the voltage stress on the fets. |
||||
| paceman Guru  Joined: 07/10/2011 Location: AustraliaPosts: 1329 |
Hi David, This would be a great project to include in the Maximite library. You've already got the code well documented and your last post explains most of the "gotchas" which you could include in a short explanatory text file to go with it (along with a link and acknowledgement to Tim Nolan's work). How about submitting it to Hugh for the library? Greg |
||||
| shoebuckle Senior Member Joined: 21/01/2012 Location: AustraliaPosts: 189 |
@crez As paceman says, when you are happy with the program and doco, you can send it to mmlib@geoffg.net and I will add it to the User Library. Cheers, Hugh |
||||
| crez Senior Member Joined: 24/10/2012 Location: AustraliaPosts: 152 |
Thanks Greg and Hugh, I am hoping to test the unit tomorrow at full current(sunlight permitting).I hope nobody is holding their breath though. I have a full time job and a hobby farm that keep me away from this fun stuff. When it's been running for a few days without problems I'm happy for it to be added to the library. David |
||||
| kebaotu Newbie Joined: 10/10/2013 Location: FrancePosts: 2 |
Can you explain the multiplier coefficients (7.5) of the INA282 to measure current? |
||||
| crez Senior Member Joined: 24/10/2012 Location: AustraliaPosts: 152 |
Kebaotu, The same line in the latest version which I will upload soon is: sol_amps = sum_sa / AVG_NUM / .001 / 60 '1mohm shunt, 60 amplification factor which equates to a factor of 16.67. 60 is the amplification of the INA282. 0.001 is the shunt resistance. At the time I uploaded the last version I was using a 2.22 milliohm shunt. David |
||||
| crez Senior Member Joined: 24/10/2012 Location: AustraliaPosts: 152 |
General Update: The hardware, in basically the form presented so far has been running fine. With 1.44 kW of panels at 24v feeding it, 200AH of batteries and a 2.4kW inverter supplying mostly daytime load, The charger hardly gets warm. The main issue with the software is a tendency to overshoot on the battery voltage. On several occasions the inverter shut down with the message "DC bus bad". I have re-written the software into an interrupt driven form and now the overshoot is much lower and no longer bothers the inverter. from watching the monitor, the overshoot is very brief. I will upload a slightly revised circuit and significantly revised software soon. Here is a screen shot of the activity of the charger over a 1 hour period: 
David |
||||
| kebaotu Newbie Joined: 10/10/2013 Location: FrancePosts: 2 |
Your system support over 30V DC voltage of the PV? I see that the capacitor is 50V voltage max support. If I put the inventor, the voltage of the battery drop. Does that affect the control of load? Your system delivers more than 40 amps? (1440W / 30V = 48ampère) You can take pictures of your circuit board? Are you using cooling part? thanks Sorry for my bad english  |
||||
| crez Senior Member Joined: 24/10/2012 Location: AustraliaPosts: 152 |
kebaotu, On a very cold morning, a 24v solar panel can go to 48v unloaded. I have tested this to 46v input. Most of the time the solar volts stay between 33v and 40v. If the system is in float mode, the output voltage will stay at the float voltage ( now set to 28.2 ) as the load changes. If there is insufficient input to run the load, the system will switch to bulk mode, and the battery voltage will start to drop. It is designed to handle 50 amps. At the moment I am only measuring input current, but the output current will be higher than the input. Q6 and all the FETs except Q5 are mounted on a heat sink that measures 160mm * 120mm with 45mm fins. It could be 40% smaller if air can flow freely. My circuit is very ugly. I have been too embarrassed to post pictures. I will put some up soon.
The inductor I1 needs special attention. I will post details with pictures. mon français est pire que votre anglais.  (thanks to google translate) |
||||
| crez Senior Member Joined: 24/10/2012 Location: AustraliaPosts: 152 |
The inductor I1 I made uses three of these: iron powder core They are stacked together. The winding consists of 18 turns of 7 strands of 1.2mm enamelled copper wire. The strands are loosely bunched together and then wound onto the three cores, pulling tight on each turn. The inductance at low current reads more like 200uH, but at high current magnetic saturation of the core reduces the actual inductance down towards 100uH. This may be a bit overdesigned, but the cores degrade if run at too high a temperature. As they degrade, they generate more heat and you end with a thermal runaway situation. I used a design tool from micrometals which gave me a good idea of what size core and wire to use . The agent for Micrometals cores in melbourne is Magcore. 
|
||||
Bryan1 Guru Joined: 22/02/2006 Location: AustraliaPosts: 1344 |
Hi Crez, YAY  my futurlec order finally arrived with the IR2184S chips and the mosfets, the shunt chip I got from TI as a sample so now I'm ready to give this project a real go. I'm now just waiting on those linear bearings to arrive so I can beef up my Z axis on my cnc then get on and design a pcb for this project.
Now the code you provided on this thread is the the latest code ? I have a couple of 175 watt 24 volt panels in parallel hooked up directly to my batterybank and I will use those to test this mttp project. Now once I can get it all working I will have a go at expanding the project so it can monitor the output power aswell and do some logging. Regards Bryan |
||||
| crez Senior Member Joined: 24/10/2012 Location: AustraliaPosts: 152 |
Bryan, I have uploaded the latest software and circuit. I have 8 180watt 24v panels on mine. If you are staying with a smaller system you could reduce the number of fets/turns on core/core sections. Mine has been running fine and its time I mounted it in a proper box. 2013-12-04_110020_MPPT24B_2.zip David |
||||
 Print this page |
