Jump to content


  • Posts

  • Joined

  • Last visited

  • Days Won


Everything posted by P1000

  1. Fluid in the container will make no difference to it surviving or not (in this case). What caused it to fail is the pressure from outside. Placing it under vacuum means pressure from outside, what's inside only matters if it lends support. In this case water will give you very little support at the bottom, none at the top of the water. Vacuum can be very dangerous, so it might actually be good that you had this experience early without any injuries. Ideally only use polymers that will not shatter like that - like polycarb instead of perspex. An aluminium pressure cooker works well if you have a suitable plate+seal to replace the lid with. Looking forward to the next update!
  2. Chopping sine waves will not work well with any inverter that I know of. I would not recommend going this route. It will work if you only use it while grid-tied, but I still don't recommend it. If you do decide to go this route - use one that uses a triac for control. If it is just a thyristor, the output will be chopped DC. Well more like chopped rectified AC. That is not a good idea for a geyser element.
  3. If they are the next best thing, you might as well buy LiFePo4 now, and by the time you have to replace them, the LTO price would have dropped so much that it works out cheaper. With batteries that have a longer life, the risk of them failing due to something else also increases a lot (longer life = more time to develop another issue). Most of the failures of LiFePo4 you hear about is not related to end-of-life, so with that in mind, I would be hesitant to spend 3x the money on something that might last longer.
  4. I think the heatbed is triac controlled - either by chopping the wave like a dimmer, or by simply switching on and off at a fixed interval (or not - like dithering with Bresenham's line algorithm). The better method to control a heater is to use Bresenham's algorithm slaved to the line frequency. That would mean that it can only turn on a complete cycle but also that it could turn on for a single cycle. All of the above methods will cause interference with inverters. If it really bothers you, the best method in my opinion would be to get a DC heat bed that runs off a SMPS at high frequency PWM.
  5. If I read that graph correctly, the 31,4A is at 48V, so it should not cause an issue. Your microwave did not consume 7222VA.
  6. I don't think there are official distributors, but I have seen individual imports before.
  7. Well, I think the advice from Coulomb is worth a lot more than my input, so you really should be thanking him
  8. Just to chime in. PWM is always inferior to MPPT, except in perhaps a very specific case, and will only be so for a small part of the year (or day). PWM has no storage component - it really is just what it says PWM. It will connect the solar panel to the battery with a mosfet until the battery voltage exceeds a setpoint and then disconnect until it goes below setpoint minus hysteresis. So it also cannot output more current than the panel current. The big loss in efficiency is the fact that you lose all the power in the voltage mismatch between panel and battery. So if you have a panel that delivers 20V and 5A (at peak power in this example case), and a 12V battery, you will only be able to deliver [email protected] to the battery, which means you lost out on 40W.
  9. All the appliances you mention work on the basis of switching an element/magnetron on or off. Lower settings simply mean that the on period will be a smaller percentage of the time. This means that the inverter should still be capable of carrying the full load.
  10. It is usually a lot cheaper to add more PV panels than erect a tracker.
  11. This should last you a while: (mine has been running flawlessly for 7 years...) https://www.sustainable.co.za/laing-ecocirc-d5-38-700b-circulation-pump.html
  12. Yes it is possible. I do not know of any like this available at the moment, though. There are a number of inverters that work like this: Battery -> boost -> HV DC bus Panels -> boost/buck-boost/etc -> HV DC bus and then from the HV DC bus -> AC 50Hz So now you just need an extra battery <-> HV DC bus for all excess loads. (The battery will be either bidir or also have a buck from HV DC -> battery, there are other simplifications as well.) Overall it adds a lot of complexity for installers and will be a nightmare dealing with complaints etc. But certainly possible technically.
  13. In this scenario, what is the definition of a "rectifier" as you have used it here?
  14. By ADuM I mean something like this: https://za.rs-online.com/web/p/digital-isolator-ics/4967519P/ For sure arresors maybe 1ohm through-hole resistors going to these: https://za.rs-online.com/web/p/tvs-diodes/0508018/ (PESD12VL2BT,215)
  15. I like the FTDI stuff. You can get them from RS, but they are a lot more expensive than others. You will probably find what you want at Micro Robotics (PL2303). On opto-isolation: That could work but it is probably not going to be that simple - you need to arrest the surge and make sure it stops before the isolation barrier (on both sides) and have a fuse of sorts and enough distance to prevent breach. I have seen a number of very well designed isolation interfaces that failed to actually stop the surge from propagating simply because it is really difficult to do well in the real world. I like the ADuM range of digital magneto-isolators for communication, but they require power on both interfaces.
  16. Not in an official manner. The fiber converter has it's 12V running over 2 pairs, data over the other 2 - so "power over cat6" perhaps. My Ratel has no PoE output either.
  17. Quite an elaborate setup - fiber converter, mikrotik router and wifi router. Usage is in the order of 7W at the DC end IIRC. So the boost converter is quite efficient. When the batteries are charged, consumption from the wall is roughly 11W.
  18. The point I am making is that Li-Ion don't like to be kept at 100% SOC - which is why you should store them at 40-50%. SLA batteries also have a life when kept at 100% SOC, usually 18-24 months. Which is why most computer UPSs last about that long.
  19. Having the batteries float at near 100% they behave very differently. I have a ratel (415 I think) that I was expecting to be in need of new batteries by now - about 2 years of service. It still manages to carry me through about 4h15 of power. When it was new it could do just about 4h30. SLA also have a standby life - usually this figure is available in the datasheets of batteries designed for UPS use (unlike lithium). The life depends on float voltage IIRC, but it usually works out to about 18-24 months. If I had an SLA system instead of the ratel I would almost certainly have to replace the batteries by now.
  20. Not quite - they are current sources so open circuit voltage will be the higher voltage. The max power point could be harder to find with dissimilar panels, but these are so close that it is hardly worth considering. The difference is barely 1% - the only reason they are binned so close is to maximize profits.
  21. I think you mean 35mm² and 50mm² - 50mm diameter would be challenging to crimp amongst other things.
  22. If all else fails: https://www.communica.co.za/products/top-ht0650
  23. I came here to say check your non-return. As for the multi valve, you won't push out much hot water before the pressure will normalize again. The system should ideally also have an anti-syphoning loop - the return from the solar collector (hot side) should go lower than the cold outlet on the geyser near the geyser-like a U-shape. This should stop or minimize reverse syphoning even when the non-return fails.
  24. In that configuration they are not exactly paralleled - it is more like 2 separate systems, each with separate battery, PVs and loads. Here they are wired almost like they would if you and your neighbor* each had a system... * and shared the same line - so more like second neighbor over
  • Create New...