Modina

It is not possible for a 3W LED to consume 40 or 50W. These bulbs are made 100% of plastic, except for the PCB that is a single layer on an aluminum substrate to act as a heatsink for the LED chips. If the bulbs would really consume 40W the whole thing would melt. Somewhere you are making a measuring error. I don't think for garden applications one needs the full brightness of a 3W. Dimmed lights are likely to give a better ambienance. Especially if you use so many bulbs. I would suggest that you connect a single capacitor in series with ALL the LEDs in parallel. Because you use a lot of bulbs, I would try to use a motor starter capacitor. It might not work, but I think your chances on 3W LEDs is 80% that it will work. It depends on the LEDs you use. It is a very quick and simple thing to test, and you might just be pleasantly surprised. You have nothing to loose. Just make sure that the capacitor is rated for 250VAC or at least 600VDC. Once you see that it works, you might want to try using bigger or smaller capacitors to get the brightness just right for your personal preferences. Even using a 0.47uF capacitor will cause enough current to power up all your 18 LEDs at a very low level. I would not want to randomly have to replace LEDs when you use that many. Another advantage of using a capacitor is that the wiring could literally be short circuited without any ill effects. Outside wiring has a much higher chance of getting damaged.

I'm an electronic engineer. 12V is high voltage for me. 😆 Once upon a time, I designed some EEG amplifiers used for neurofeedback.... Brain waves are measured in a few uV. My bible is The Body Electric by Robert Becker. Bone has piezo-electric and semiconducting characteristics. That book is the reason why I dislike strong H & E-fields, RF radiation and that sort of thing. Maybe you can use that heater that you call a resistor for both wind gen functions. To increase the source impedance and as a real dump when the wind gets too strong or batteries are full. Might have to place a tap somewhere in the middle for MPPT usage. And the full length of the wound resistor as the dump.

May I suggest a simple test. Try and place your hand on one of those 60W bulbs and see how long you can keep your hand on them. Then repeat the test with your 3W LED. Doesn't matter what you do, at the end of the day, just about all electric power is converter into heat. Sound and light generation is a very tiny amount. Even mechanical power from an electric motor ultimately is turned into heat, with the exception of adding kinetic or potential energy in applications such as water pumping and the like. I think your assumption of a PF = 0.8 is way, way to high. BTW, these cheap LEDs are all overdriven. The manufacturers purposefully do so, so that these bulbs very seldom see more than 18 or 24 months of life. Otherwise, who would buy LEDs in future? Open the LEDs and modify them to run at half power and they will run for many years. You can check out the youtube videos from Big Clive. He has made a number of videos going into the details. See my other posts about this matter. You could also connect a 0.22 or 0.47uF high voltage cap in series, and that would also dim these lights and reduce the power further.

Can a typical Axpert inverter be used for 230VAC and solar MPPT charging only, without the DC/AC inverter being active? Say I have a 12/24/48VDC only load and do not want the expense and clutter of using a separate 230VAC battery charger and a solar MPPT charger, but I don't want the inverter running and gobble 30 - 60 Watts of power. I doubt that it is possible but I don't yet know the Axpert inverters that well. AC charging will likely not work. How about only using the MPPT?

Heine, switching high power DC is not so easy. I would not recommend an electro-magnetic relay due to DC arcing. Did you maybe see a post someone did about 2 weeks ago where he was using a total of 4 solid state relays? He included a wiring diagram. I think he had exactly your problem. Solid state relays are plentiful for AC applications. DC SSRelays are not that common and they become expensive. The problem is that for DC the switching voltage is limited. So he used two DC relays in series. I am not so sure if that is a good thing. You need to switch at EXACTLY the same time. And there is no zero-crossing point you can wait for, like in AC. At the end of the day, it might be best to invest in a separate MPPT battery charger that you can sell at a later stage again, rather than buying a lot of solid-state relays and other hardware that will not exactly be cheap. You are less likely to sell second hand relays and other bits and pieces for a decent price. However, a MPPT will not be cheap. If you do decide to use an electro-mechanical relay you will need to put a snubber across the contacts to try and limit the arcing. A snubber is a capacitor and a resistor connected in series. Obviously the relay contacts will need to be rate for the voltage and current you intend switching and allow a large safety margin. Let's see what other people have to say.

Sorry @Scorp007. I lost my mind. You are right. But what a waste of power. Something like that is a heater and shouldn't be called a resistor. SunSynk must be thinking of seriously large wind turbines. I wonder if the current through the field windings of a generator could be manipulated to vary the "apparent" output impedance. In this way no power would need to be burned. But I am no power or motor specialist. I rather shut up before making another dumbass of myself. LOL.

@Sc00bs I am allergic to WiFi. Haha. No, these smart plugs can only measure reactive power - at least as far as I know. I want to be able to measure the power factor or "real" power. If you don't know the PF, you could make large errors as the real power could be substantially lower than the reactive power. PF is a difficult subject to get your head around. It has to do with the current and voltage being in phase (in resistive loads) or out-of-phase in inductive or capacitive loads. The PF is why inverters normally have two power ratings. One expressed in KW and the other in KVA. KW is into a resistive load and is always smaller than the KVA ratting. We tend to loosely interchange the two, but there is a difference.

A three phase generator will have very little (rectified) ripple. The phases are overlapping and never reach the zero volt point in the way a single phase does. A SMPS should ideally run far over 20KHz and should be able to handle large input ripple, but it is always advisable to keep the ripple reasonably low. In a single phase application one would definitely need to place a smoothing cap after the rectifier. If there is the requirement to increase the source impedance, that would have to be done after the capacitor. A 1KW 1ohm resistor??? That can't be right. If we assume an input current of 40A (at a high voltage that would be a massive input power) a one ohm resistor would **only** dissipate 40W.

Here is an explanation of what can cause correctly functioning LEDs to flicker. (Faulty LEDs can also flicker during normal on conditions). 230VAC LED lights fall into two general groups. Those that use a so-called transformer-less power supply and those that use a proper SMPS (switched-mode power supply). The transformer-less designs use a non-polarised capacitor in series that acts as a voltage dropping element. The capacitor is followed by a bridge rectifier, a smoothing capacitor and maybe a resistor and/or zener diode (A zener is a voltage clamp). These circuits are cheap and are used for low wattage LEDs (typically anything up to 3W). These LEDs should NOT flicker. Any leakage current that might arise from capacitive cross-coupling between long cable runs or other conditions, will turn these LEDs on at reduced brightness. Sometimes so dim that you can hardly see it. Good quality LEDs and the higher powered LEDs (typically 5W and more) use a small SMPS. Here a PWM regulator chip is used to drive the LEDs at a constant current. The 230VAC is first rectified in a bridge rectifier, then smoothed by an electrolytic capacitor and then chopped in a step-down PWM regulator. The regulator IC will require a minimum operating voltage to wake up and become active. This voltage might be as low as 40V or higher than 90V. There are many different PWM regulators on the market. Each one will have it's own characteristics. So here is what can happen. The leakage current will find it's way through the rectifier and slowly charge the smoothing capacitor. The voltage over this cap will gradually rise. At some point the voltage will have reached the threshold where the PWM regulator chip will wake up. The chip will now attempt to turn on the LED at it's designed brightness/current. The smoothing capacitor will now be discharged near instantaneously. The voltage will drop below threshold and the PWM regulator chip will shut down. The leakage current will again start charging the capacitor and the cycle will repeat. To prevent this, one could connect a resistor across the LED, but this will waste some power when the LED is on. Alternatively, a capacitor can be wired across the LED. This will also bleed away the voltage. The advantage is that the cap will not consume (real) power when the LED is turned on. I have found certain PWM chips that will not cause flicker and the bulb will sometimes behave in a similar way as a transformer-less LED bulb. It all depends on the PWM regulator's internal design and component values of the circuit. I would be surprised to find a Osram or Philips LED flickering. It's the Chinese manufactures that want to save every last cent and take shortcuts...

@Coulomb Yes I agree. In theory one would be able to add a series resistor after the bridge rectify but this might have to be a very high wattage resistor mounted on a heatsink. Not ideal for efficient power transfer. I need to see if I can calculate the internal resistance of a typical panel to see where one would end up. Obviously whatever the resistance, one would need to multiply this by a number of "panels" that would typically reach the 320VDC level. Adding a series resistor would be alright for a 100 or 200W system. If one talks about 1KW+ it gets difficult. For me, any power manipulation in the KW range is "difficult". The principals are the same. The circuits are the same. It's just that the holes in the semiconductors, the holes in the pocket and the brown marks in the undies are potentially much bigger. I would have expected such hydro-systems to use a 3-phase generator. I believe that most wind generators use 3-phase. Converting that to DC would obviously reduce the ripple voltage. I do not follow your logic about needing a circuit that is similar to a motor controller A MPPT is basically a SMPS that has an integrated (firmware controlled) battery charger. So I don't see what the problem would be to feed the MPPT hardware with a constant voltage and manipulate the battery charge current not only for battery safe operating area but ALSO for limited input power (so as to not overload the generator). I am under the impression that one should be able to re-configured it for such constant DC voltage with firmware only, and retain the hardware architecture. In other words, use the MPPT hardware to implement a "normal" battery charger. This would only be possible for someone that has the source code - i.e. the manufacturer. Perhaps this would be redundant on inverters with only one MPPT, but that sort configuration choice should at least be made available to inverters that have more than 1 MPPT.

Thanks @Bobster. Yes those are the same functions that mine covers. So if I can't fix mine I will buy a new one. Sometimes new models are watered down in functionality. One never knows. Just wanted to make sure. For me the PF measure is important and few meters support it. Technology life-time follows the so-called bath-tube curve. High failure rates at the start (due to manufacturer defects), then a long period of rock solid performance and then after a certain period (often designed into the product) a high rise in failure.

So your meter wasn't very safe and effective and died suddenly? Mmm. I have a Ellies FSEMS efergy meter that is a few years old. It looks identical to your Kill-a-dud. Mine became erratic and I then noticed that the internal battery was leaky and had died. I couldn't get the right replacement and decided to use a super-cap. Now the cap seems to have charged beyond the intended voltage. Stupid me, I should have expected that. I need to spend more time on it and see if I can get it going again. My efergy meter could measure the Power Factor (PF) - Has the kill-a-watt meter also got that function? I really would like a meter that can measure the PF. If you are not interested in the PF, you can get one of these: https://www.sonoffafrica.co.za/product/sonoff-powr316/

You don't need a 50A rating on the bridge rectifier. 25A would work well. They are so cheap that I wouldn't see the need to go lower than 25A. I like good safety margins.

Other people will chip in, but here is my take on things. The hydro generator will have the same problem a petrol generator has. An unstable 50Hz frequency and fluctuating voltage. It is ironic that most loads would tolerate that (within reason) but an Axpert and many other type of inverters will not. Sad but true. My believe is that you could run the 230V through a full-wave bridge rectifier followed by a electrolytic smoothing capacitor. This would give you about 320V DC. This voltage should sit bang in the centre of a high voltage MPPT. Your 1000 odd watts should be well within the power/input current limit of an Axpert with a high voltage MPPT. @Coulomb will be very quick to warn us that you cannot do so for high power generators (say 5KVA) as they would be able to source too much current for the MPPT to handle. I am sure he is right, but I am still scratching my head as to why a very intelligent DSP that is in charge of the MPPT, is not clever enough to self-limit the incoming power. If the above is not possible, the only other solution would be to bypass the Axpert and charge the battery with a dedicated battery charger. Getting a 1KW battery charger would be hugely expensive. PS. A 50A bridge might cost you R 150. And a high voltage capacitor a bit more. So this could be implemented at a very reasonable cost. You would need to mount a heatsink onto the bridge rectifier.

Samsung normally quotes maximum and average consumptions on TVs. You can expect ANYTHING from no-name Chinese imports. They will LITERALY produce and sell stuff that just cannot work, ever. Youtuber Big Clive has shown many such products. If mains powered, these can also be very dangerous in terms of personal safety and fire. If it comes to branded products you can be assured that if the sticker says max power 70W that the item will ALWAYS remain under that threshold. The German VDE, Euro CE and US standards are very strict and a manufacturer would be in big trouble if found to be less than truthful. In fact, if an electronic gadget is rated at say 40W max, real-world power consumption is likely to be much lower. Not so for electric high power devices such as irons, kettles, etc. I don't know of any 1.2KW kettles. Users would complain. Complain why their kettle is so slow. Perhaps you get 1.2KW kettles for some camper type of application. We are not so fortunate as the Americans. They can order a 230V kettle form overseas and it will produce half the power when connected to 115VAC.

Fridges are a problem for small inverters. I have seen youtubers claim that they run fridges on modified sinewaves but my advice is not even to try doing so. They might run, but not for very long. I suspect the compressor would run very rough and would likely overheat due to all the high frequency harmonic content. I would go further and say that even pure sinewave inverters rated at or below 500W might not be able to start a fridge. The inrush current is just so high that the inverter is likely to shut down. I am sure a 500W Victron will have no problems at all, but a 500W Chinese inverter is probably rated in ponies rather than horses. It is likely to go on strike if asked to deal with a fridge's start-up conditions. There are exceptions and in this case its inverter fridges. They normally use 3-phase motor inverters that have soft start functions as well as speed control. These fridges should be OK running from modified sinewave inverters and have very little inrush current. LG and Samsung sell inverter fridges on the high end at around R18K and upwards. Inverter fridges are claimed to have higher efficiency. My recently purchased Defy "solar hybrid" is a small 157L unit and sells for around 6K. This has a Danfoss-like inverter. It has a very slow power ramp-up that takes several seconds. So no discernable inrush current.

@Bobster. 170W for a 40" TV is exceptionally high, I don't think that is for an LED TV. It's what I would expect from LCD cold cathode backlight technology. What make of inverter does your mom use that is so low on losses? Many inverters do not specify the efficiency and even less, give the no-load quiescent current. The Victrons seem class leaders in this regard and I have seen one or two others. There is junk out there that will have 40% losses. Those can be used to keep your coffee cup warm.

Of course you can run a TV from a generator. A generator will run nearly anything because it generates a true sinewave. What I meant is that a generator is not practical if all you want to do is watch TV. If you insist watching on a 55" screen then you might have to use a generator. My point is use a modern, smaller sized TV which will not be a high load to a battery system. Battery and solar systems are so expensive, that if your TV is too power hungry, it would make a lot of sense to replace it. So doing will reduce the battery/inverter size. You spend much less on a new TV than to power an inefficient TV from lead acid batteries that will need regular replacement. Many years ago I brought a 2.5KVA generator. It had a carburetor problem right from the beginning. It ran too rich, had high consumption and wouldn't start easily. Later I brought a 5KVA. It leaked oil. I never used it. Good luck to anyone who thinks he can buy a cheap, made in China, genie and use it EVERY DAY to watch TV. If you do that, you will either need to fiddle with it like a lawnmower, or you will replace it faster than a lead-acid battery.

A generator is NOT the right thing for powering a TV and lights. But if your dad say, wants to use power tools then it's perfect. I rather replace a R2500 battery once a year then schlep around petrol every third day, listen to the racket, having to manually start the genie, twiddle with the choke, tweak it's throttle and inhale stinking exhausts. In an emergency, I prefer another 4-wheeled genie. It's called a car - yes it only has 12V and limited power. At least it is silent, clean burning and I would expect it to not be heavier on fuel than a primitive, carburettor-fed genie.

@Scorp007 Those UPS systems tend to run very hot. At my previous company we started cutting holes for cables and placing the battery next to the UPS on the outside, to increase the MTBF of the batteries. SLA are not "sealed". I have sometimes "repaired" SLA batteries by simply popping off the top plastic lid, removing the 6 rubber grommets and filling with distilled water. Then replacing the grommets and giving them a good charge. SLA's are not wet batteries but still hate heat and dry out. I could use "repaired" batteries to operate gate motors without problems for many more months. Heat and vibration are the two killers of lead acids. Many BMWs and some other cars, had the battery in the trunk. Less heat and less vibration. They typically last 5+ years. I think certain BMW 5s had the battery up front, and people where lucky if these survived 24 months - modern engines are covered in plastic covers and trap the heat. @fuzex For any TV run from batteries, I would suggest the following: 1. Make sure it is a LED TV and not an old LCD - LEDs use about 1/3 of a LCD TV 2. Try to limit the screen size of the TV. Large screens = lots of power drawn. For R3000 you can buy a new 32" Samsung. A bargain considering what batteries cost. Your batteries will thank you. For the small scale user only wanting a TV, WiFi & lights, if he is a bit DIY inclined, I would suggest getting a TV that runs from a 19VDC laptop-type power brick. Many of the 32" Samsungs use this setup. Then get a DC-DC boost converter for about R120 and run TV, lights and WiFi straight from 12V. Advantages are: Safety (no high voltages that are unlikely to have earth leakage protection in these small systems), no noisy fans, no inverter expense (you still need a charger) and most important of all, no power losses due to the inverter. These losses can be substantial. I have a 24" Sinotec TV which I brought a few years ago specifically for load shedding. It runs straight from 12V and consumes about 1Amp when set to a medium brightness. i.e. 12W. A modern 32" LED TV should be about 20-24W. An old 32" LCD will use over 70W by comparison. Large screen LED TVs will likely also use 70W and more. Similar logic can be applied to PC monitors. Desktop PCs are not power optimised and are heavy on juice. I personally have a dislike for laptops and prefer a desktop. When replacing a desktop, consider one of these mini-PCs (the size of a paving brick, such as the Intel NUCs). They run much more efficient and are available from a Celeron to i7. So although small, they have a lot of CPU power. And again, these are powered by external 19VDC laptop-like supplies.

@MakoThat's nothing. A few years ago I sold a house in Pretoria and the electrician that came to do the COC didn't even have a multimeter. But I got my COC.

@Steve87 Yes I agree with you. So if I understand you correctly, there is no extra training requirement for a master electrician to legally issue solar installation COCs? Just a reluctance of some to do so?

All I can say to this subject is that if your electrician is required to perform earth measurements for the purposes of issuing a COC, I would recommend doing so after you had some good soaking rain. That should make the figures pass with flying colours. Otherwise you might need some very long earth spikes. Not so?

Only a master electrician can issue a house COC. To issue an electric fence COC the electrician needs to have done an extra exam. Can any master electrician issue a solar COC or is he also required to do an extra course/exam? COCs where normally filled in by hand on a pink (or was it green) form. So @Peter V it sounds as if the system changed. Has the content of the COC changed as well, or has the form just moved into the digital domain? It would be good if someone could post a copy of this COC "solar supplement" so that people still on their way of getting their own solar system could view the technical aspects that need to be signed off. I also have questions around SANS. What document (number) covers single phase residential wiring? Is only one document required, or are there multiple documents? For instance, does it cover DC and low voltage DC regulations as well, or would one find that in a separate document? Finally, what would it cost to obtain the required document(s)?

Yes, I was told by technicians how they as apprentices charged capacitors to a few hundred volts and then threw them around, shouting catch.