 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 : Solar : Solar Hot water system (thanks to forum, Warpspeed and others)
  Page 3 of 3 |
|||||
| Author | Message | ||||
Murphy's friend Guru  Joined: 04/10/2019 Location: AustraliaPosts: 671 |
Mike, I have re wound quite a few toroids, see also my old 'tinker' threads. This was actually quite enjoyable when doing it the easy way. If you read many of my posts you will see it's a learning curve that I, like everybody else went through. May I suggest you forget about the three stack idea. With a tall narrow 'hole' it's difficult to keep the wire flat against the hole wall as you must to avoid running out of 'hole' space for your primary. A two stack from 2 x 3KW Aerosharp cores is as much inverter as you ever want, it can start and run amazing loads. These home built inverters have come a long way since 'Oztules' early Power Jack experiments. I have gone along with the new ideas as they came up, the early inverters had a nasty habit of blowing Mosfets for no apparent reasons. My inverters have now all been updated to the opto isolated version, this has proven soo much more reliable than anything that came before. Have fun. |
||||
KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1859 |
Yes I've read every build post (including your posts) to see what hurdles were overcome by everyone. My main problem is my age and recovering from another recent health setback, it means that doing things like this are a bit tougher. Fortunately I have the technical skills. I'm going with the OzInverter. I was lucky enough to find one 2.?kW AeroSharp. But I'm using 2 or 3 toriods from the small SMA GTI. The AeroSharp is a similar size but much taller, so 3 x SMA = 2 X AeroSharp. I plan on starting a build thread (or is might be a help thread?) if I have problems. Just testing the caps and already found issues with one under soak test. If that one had been used in an inverter, it would have eventually destroyed the inverter with any heat built up around the CAP. Next is testing every MOSFET, fortunately I have all the test gear. Looking forward to the adventure. Mike. Edited 2022-12-24 16:59 by KeepIS NANO Inverter: Full download - Only Hex Ver 8.1Ks |
||||
| Murphy's friend Guru Joined: 04/10/2019 Location: AustraliaPosts: 671 |
Age isn't the problem Mike (I'm 77) but hand and arm strength is. You might get a little fitter winding a decent size toroid. Just take your time. I did all my winding sitting down with plenty of breaks to admire my handiwork  .I would not wind a stacked core if the blank hole was smaller than 90mm. 100mm is much better, especially if you plan to use heavy insulated welding cable for the primary. Good idea to test the caps, I did leakage tests at their rated voltage and discarded those which were too bad for my liking. Biggest problem with the 'ozinverter' design is the power board layout, it must have plenty of cooling if one runs it on high power. |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1859 |
Thanks, I see you found the new thread on the inverter that I started. Fortunately I'm not trying to make it fit into a small cabinet. If I go with the OZ I was planing on changing the layout to make it easier to get at everything, so heat sinking and cap cooling will be revised. Heat is CAP cancer and rapidly destroys CAP life expectancy. Catch you on the other thread. Really dull overcast yesterday with drizzling rain, took until 4.30pm for the controller to switch off the HWS thermostat, but it did. NANO Inverter: Full download - Only Hex Ver 8.1Ks |
||||
| Phil23 Guru Joined: 27/03/2016 Location: AustraliaPosts: 1667 |
Hi All, Dropped in this morning with a very similar concept in mind. Presently my Hybrid System is running well under capacity. Economizing usage has me in a situation where I have at time 1000 to 1500 of Load capacity going unused. The thought of dropping our HWS off the Ripple Timer control came to mind. Basically, a decent relay/contactor to switch across to the inverter output. Primary issue though is that it's a 310 Litre HWS with a 3600 element. I'd be wanting to load the inverter down with basically somewhere between 500 & 1500W (AC) depending on conditions. First thought that comes to mind is SCR/Triac type control like a light dimmer, but have in the back of my mind the fact that my variable heat gun does funny things when I lower it's output when it using inverter power. Any thought's or previous experience on this approach? Cheers Phil. |
||||
phil99 Guru Joined: 11/02/2018 Location: AustraliaPosts: 2557 |
Does your heat gun use phase control or burst control? Burst control typically switches the element on / of for up to 10 cycles at a time. Jumping between 0 and max. all the time might be what gives your inverter a hard time. A Triac triggering after 90 deg. in each half cycle might be less demanding as the inverter never has to supply the full power. |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1859 |
There was a discussion somewhere on the worst loads for an inverter (especially with a big Toriod inverter) and some of the crude methods used to change power levels. My ancient ARLEC heat gun (1.7kW / 2.4kW) causes no issues with the HV (no transformer) pure sine wave inverter. Don't know what it uses to select HI/LO pwr. The Solar HW controller in this tread is obviously DC in and out and in my case, setup as pulsed DC at all times to keep the loaded output around the MMP of the panels. NANO Inverter: Full download - Only Hex Ver 8.1Ks |
||||
| nickskethisniks Guru Joined: 17/10/2017 Location: BelgiumPosts: 462 |
Phil23 Another solution is make use of a transformer, an autotransformer or rheotor/variac. Or ading An extra heating element. |
||||
| Phil23 Guru Joined: 27/03/2016 Location: AustraliaPosts: 1667 |
Don't know, the heat gun may be an old Ryobi. 2nd element would be nice, but the tank has no facility. So probably just need a Triac based control that can handle potentially 15A. Would need to be PWM controlled so I can drive it with an ESP for control by Home Assistant. |
||||
Revlac Guru Joined: 31/12/2016 Location: AustraliaPosts: 1140 |
Phill23, I have used one of those "10000W 25-A High-Power SCR Speed Controller Voltage Regulator Dimmer Thermostat" for soft starting large brush motors, I haven't put a scope on the inverter to see any adverse affects. Probably should start a new topic for this. Cheers Aaron Off The Grid |
||||
| bpjh56 Newbie  Joined: 04/06/2025 Location: AustraliaPosts: 2 |
Hi, I think I understand the circuit operation but I'm lost on a couple of things. Firstly, the IGBT used appears to be a IKW30N60T. It's a bit hard to see as the backshed id obscures some of the markings in the schematic. If it is this device, the safe operating area graph on the datasheet gives the maximum current at 200V to be about 0.9A. The 3 IGBTs in parallel gives a safe current of 2.7A. In a post earlier, the average power was given as 1.47KW with an average voltage of 187V, giving a current of 1470\187 or 7.8A. I'm assuming here that a pulse length of 98ms is close to DC. Where am I going wrong? Why don't the IGBT's cook? Another question I have is with the capacitor in parallel with the solar panels. The original author, warpspeed, indicates that the size of this capacitor is flexible. So why not have a much smaller, and cheaper, capacitor? Like the units that are used for motor start applications? It seems frequencies of up to 2MHz are OK with the LM555 timer. However the FOD3182 has a maximum switching speed of 250KHz. Even at 200KHz the discharge pulse becomes 5us, raising the safe current maximum of the IKW30N60T to 80A per device. I assume the role of the capacitor across the solar panels is to provide a load that pulls the output voltage down to maximum power point.So you don't want the capacitor to fully discharge either as you would have effectively a short across the solar panel. So how much should the capacitor be discharged? My heater has a resistance of 15R (ohms). If the capacitor is discharged by one third in each 5us pulse, the capacitor size to do this is 0.33uF. Could someone enlighten me as to where I'm going wrong? Thanks in advance, Brendan |
||||
| phil99 Guru Joined: 11/02/2018 Location: AustraliaPosts: 2557 |
That is in linear mode where the IGBT has 200V AND 0.9A. As it is switch mode the IGBT gets 187V OR 7.8A. When it is off there is no current so no heat, when it is on almost all of the voltage is dropped across the heater and almost 0 across the IGBT. The heater gets hot and the IGBT doesn't. During switching the IGBT is in linear mode which does heat it somewhat. The less often it switches the less wasted heat and stress on the IGBT. With a low switching rate an oversize capacitor keeps the panel voltage stable. |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1859 |
phil99 has eloquently answered the IGBT question, and most of your questions are answered in the thread. The capacitor bank, charged by Solar, is the power source for the heater, pulse power discharge of a 5000uf capacitor into an 18Ω HEAT-BANK, which is what an insulated HWS is, in good solar that switches at between 10hz to 45hz. In my old system of panels, PWM was charging caps to 220v and discharging to around 155v, as there is always PWM switching, the existing thermostat in the hot water system was used without any problems. (Obviously the thermo must never switch DC) However I no longer use this. The solar panels are now connected to a BIG off Grid inverter, the SMALL size HWS works beautifully in this situation, it's run time is around 10 minutes at 2.8kW with normal usage and a run time of 40 minutes with long showering (my Wife) at night in winter. If you have small Solar and/or Battery power, then set it to disable the HWS at low battery state. This is better utilization of the solar panels and more efficient use of solar "most of the time" !! Obviously if you are just trying to run a small HWS off grid with a few panels then the PWM (cap charge) controller works pretty well, no "LONG" showers at night though, unless you like warm (not hot) water in the morning. _ Edited 2025-06-07 10:06 by KeepIS NANO Inverter: Full download - Only Hex Ver 8.1Ks |
||||
| bpjh56 Newbie Joined: 04/06/2025 Location: AustraliaPosts: 2 |
Thanks for the replies. Appreciate the time you both have taken. @phil99. From the scope shots KissIS posted, it appears that the IGBT conducts for between 20ms and 50ms with an equal off time. The safe operating area graph for the IKW30N60T is figure 2 on the datasheet. I used the DC graph as it was the closest to the operating conditions of the heater. The DC graph, as I understand it, is not operation in linear mode. The device is being driven fully on. The next graph in Figure 2 is for a 10ms pulse. I can't find what is the off time in the datasheet. So to be pedantic I should interpolate between DC and the 10ms pulse graphs. It makes little difference though. At 10ms, the maximum current is about 1A. I don't understand how you get a rating of 7.8A at 187V. This implies a pulse length of something between 50us and 1ms according to the SOA graph. I agree that there is a switching loss each time the IGBT is turned on or off. This is reduced by decreasing the time needed to charge capacitance of the IGBT and is the role of the FOD3182. So why the 10R resistors in the gate circuit? Would they not simply increase the time needed to charge the gate capacitance and increase the switching loss? If the FOD3182 drives the IGBT fully on the most significant loss would be the roughly 2V drop across the collector-emitter junction of the IGBT. If each IGBT carries 2.5A each, it would dissipate 5W. So it seems to me that using a smaller capacitor and switching it faster should work. I'm looking for a second hand capacitor but if I can't find one, motor start capacitors are about $40 against $150 for a 4700uF part. @KeepIS. The depth of the capacitor discharge was in the thread. From your scope shots and measurements it about a third. I have a second hand 135L water tank and about 3KW of panels but no inverter. I've plumbed the electric water heater into the water inlet of a gas hot water heater which feeds the house. The idea is to preheat the water before the gas unit and save on gas. Hence the interest in this circuit. It's obvious that the circuit works. I just wish I knew how it works so I could vary the components if need be. Thanks again to both of you. |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 1859 |
That 4700µf is made up of scrapped electrolytic caps from old Gridtie inverters from scrap merchants of other sources. Why not test your idea with a single panel and a resistive load with temp sensor. Keeping in mind that this controller works due to the nature of an insulated water tank retaining most of the power pulse as heat as the water temperature slowly rises with each pulse. Very fast PWM frequency will likely mean using a tank temperature sensor to cut the PWM off. Cheers. NANO Inverter: Full download - Only Hex Ver 8.1Ks |
||||
| phil99 Guru Joined: 11/02/2018 Location: AustraliaPosts: 2557 |
An important point. At high frequency the Off time will be too short to extinguish the arc when the thermostat opens. The gate resistor value is too low to have much effect on switching time but should be high enough to dampen resonance between the gate capacitance and PCB track inductance. Without it there is a chance the IGBT will go into RF oscillation and overheat. Data sheet pulse tests are usually done with a high Vce. Here Vce is less than 2V. The frequency is high enough to average the temperature rise so IGBT dissipation needs to be multiplied by the PWM ratio. Eg 50% on time for 5W gives 2.5W. |
||||
| Godoh Guru Joined: 26/09/2020 Location: AustraliaPosts: 515 |
I have a similar setup as KeepIs. i have an 80 litre hot water tank and a small solar setup on a shed. I changed the element in the tank to a 750 watt element. And as the small system has a 24 volt 100 amphour battery bank with an mppt controller I just use a combination of a Timer to bring the unit on between the best hours of sunlight that the panels get, and a voltage controlled relay that switches a solid state relay on to power the element. It works pretty well, and I have hot water to wash my hands after working in the shed. Pete |
||||
| Page 3 of 3 |
|||||
 Print this page |
 |
The Back Shed's forum code is written, and hosted, in Australia. | © JAQ Software 2025 |
