Jump to content
Genetry Solar Forums


  • Posts

  • Joined

  • Last visited

  • Days Won


Everything posted by kazetsukai

  1. I set off speed to 1% to keep fans from spinning up and shutting off rapidly at night. Kinda like that, but would be nice if I could make this figure a little slower. Other ideas for fan control.... load offset, temp control. Fan speed (percent) = Math.min(100, thermistor controlled value + (load pct - load offset) * load factor) or some fancy algorithm like one-way "inertia" (fan speed changes instantly going up but very gradually while going down). In all likelyhood just about any "algorithm" suggested, even if it sounds like it makes sense, will not in practice. 😃
  2. To clarify: adb shell allows one to get to a terminal on a locally connected android device, provided they have debugging enabled on the device.
  3. I looked into the issue as well. Are you familiar with adb? A simple test: > adb shell flame:/ $ ping genetry-34febd9e.local ping: unknown host genetry-34febd9e.local 2|flame:/ $ ping genetry-34febd9e PING genetry-34febd9e.lan ( 56(84) bytes of data. 64 bytes from Genetry-34FEBD9E.lan ( icmp_seq=1 ttl=255 time=74.2 ms 64 bytes from Genetry-34FEBD9E.lan ( icmp_seq=2 ttl=255 time=93.0 ms 64 bytes from Genetry-34FEBD9E.lan ( icmp_seq=3 ttl=255 time=114 ms On Android, using http://genetry-34febd9e.lan I'm able to talk to the inverter, both in firefox and chrome.
  4. MDNS resolution on desktop Linux takes place at NSS (Name Switch Service), and is configurable. Browser really shouldn't matter. IIRC ".local" is just a Linux desktop convention, which hints "please attempt to resolve this hostname via mdns", Android doesn't seem to share this convention. The inverter should still be accessible using MDNS on Android by omitting .local at the end of the address... I've also seen .lan among other things.
  5. Yesterday I ran the above oven for a little over two hours, no hiccups at all. Knowing virtually nothing about how the fans are programmmed, behavior seems like it could be better- I would see fans spin down when load was high (oven elements switch on/off based on internal oven temp) only to spin back up when temps rose again. Also, when the food was finished cooking the fans continually readjust while the inverter as a whole cools down. I think cooling back down to average temperature should be done more consistent and aggressively, then shut the fans down to their minimums rather then modulate down. Ultimately keeping the inverter cool is the highest priority... its just more "user friendly" if the fans adjust as infrequently and gradually as possible. Just an idea! Anyway I knew the already warm cabinet would probably not be a great environment for the inverter, so yesterday I installed an exhaust vent in the cabinet. Air at the top of the cabinet is sucked directly outside. (image of cabinet with vent fan) ( Now this creates negative air pressure in the cabin, causing air from anywhere it can enter to enter (mostly at the front door), but I think this is a win compared to the alternative of much hotter air in the cabinet leaking out into our hallway. The air blowing out does feel hotter than the summer ambient air outside. The blower fan is thermostat controlled, so it will throttle itself as needed. Its silent on the highest setting... I have another one of these fans below in the battery compartment. During winter this vent will be capped and not operate, as will the battery box vent.
  6. Well, thanks to help from @Sid Genetry Solar with the inverter the past couple days, I got it running yesterday. He also resolved the charge/ATS issue for me, so I'll be gearing up to get that all installed here shortly. Since yesterday evening the inverter has been humming away in its new little home: All looked well: Shortly after I calibrated it, switched bus power over to the inverter and went to bed. I set it so the fans would stay on low rather than spin up every so often, made it easier to sleep. I wonder if I could get the fan speeds a little lower even for the "fan off" setting. Today without realizing it I had kicked off a mini stress test when I went about my business and threw a pizza in the oven! 120V Ninja convection oven, pulls around 1800W-2000W. Remote display is such a cool feature for times like these... Solar input jumped with the new load: And my yield today has been a bit higher after running the AC all night! Looks like I have a solid 6-7 days of standby power for extended overcast with this new battery bank. Overall super excited to see this all come to fruition. Its been months in the making. Thank you Genetry Solar! *munches on solar cooked pizza*
  7. Sorry, to clarify: The end user wouldn't do this, you would use the Wiki as a knowledgebase, and occasionally also to compile PDF documents from several articles in manual form, hosting that PDF much like you do today.
  8. Well, examples... here's an article showing the some of the surface potential of print media CSS: https://sympli.io/blog/a-quick-guide-to-css-for-printable-webpages/ Idk what distro you use, I run Arch linux. Both MediaWiki and DokuWiki is in the official repos: https://archlinux.org/packages/?name=dokuwiki https://archlinux.org/packages/?name=mediawiki And have articles for setting up: https://wiki.archlinux.org/title/MediaWiki https://wiki.archlinux.org/title/DokuWiki Docker compose is probably the fastest way to get this up on a Linux box. https://hub.docker.com/_/mediawiki/ Again I'm happy to build out a demo of this if you'd like... I've got a week or so of free time before I start a new job.
  9. I was thinking print-media CSS (CSS that applies to web pages only during printing) could be used to convert an article/series of articles into a PDF document equivalent to that of a manual, using the "print as PDF" feature common in modern browsers. I'd think that modern open source wiki software has decent default print-media CSS, but it could be tweaked further to achieve a desired aesthetic. If you're ever interested in giving it a whirl, I could setup a demo MediaWiki or DokuWiki, copy some of the current manual content into an article, and see how it "prints".
  10. I had the chance to read through the manual today and found a few sections citing things like: "If you have this revision then this will be the case, else that will be the case"- examples being the gen start amperage/connector and the "critical warning for AC input". No complaints about the manual- its very thorough, reads well, aesthetically pleasing as far as manuals go. Going to be my go-to for questions about how to do X or Y. This is purely a look at saving time long term maintaining documents for each model/revision... a Wiki might be easier maintained over a longer term than a single document- You could have a heirarchy/page flow like: Genetry Solar (general page) PJ Inverters (list of) GS Inverters (list of) 6000W Revisions (and differences) 3 phase wiring Daisy chaining etc 3000W Revisions (and differences) etc Determining which revision GS inverter Common Inverter Info Wiring a split phase inverter Wiring a single phase inverter Determining which inverter you have Wifi Board Setup Wifi Updating Versions / Changelog etc One can link between these pages to reduce redundancy, and most Wiki software also have a form of VCS built in. Some may even be suitable as a source to generate manuals by cherry picking articles to use as sections. On the flip side, maybe manuals and wikis are separate concerns- product manuals do not equal knowledge bases for instance. Thoughts?
  11. Nope, something is wrong. After a while it has a fast blink red with the beeeeeeep.
  12. Oh, it came on again this time after letting it sit a while. Having the BMS cut out power on it like that must have put it into a weird state.
  13. @Sid Genetry SolarI got it hooked up, powered on. Fans spin up hiGH then go silent. But as I was playign with the BMS (temp sensor) one of the times the inverter came back on, it came on with a red light + beeeeeeeep. I couldn't see the screen, don't remember if it was blank or what. Turned it off and waited a little while... now I hit the button, it turns green and no activity. Uh oh... No smell or anything, just silent with a green light. No WiFi board screen.
  14. Just installed my new LFP bank which gets similar figures. Before this the battery bank was too small to ever get near those kind of daily numbers... What's your system like, if you don't mind me asking? I'm in a bus conversion, so similar use case to that of an RV. The talk of Rev. C boards potentially having dual input voltage support really has my attention...
  15. In light of some conversations around 240V charging problems, I've decided to hook up the inverter as I had my PJ inverter- output only to a manual changeover switch, no input/ATS/charge. ATS/charge were two of my most anticipated features- the cool as heck ones I've been waiting for. Just plug in an the inverter would switch over and charge from shore- sweet! However, the lack of a double-pole relay internally to disconnect L2 causes issues with my use case. My use case means I cannot avoid a ground/neutral bond upstream of the input, and I must assume ground is shared between N-input and N-output as I only have one ground available to me, I need a ground connection to chassis, etc. In retrospect I should have given 120V charge a little more consideration (for other reasons, primarily generators) but I do believe 240V input is the way to go for me. Every outlet in my bus is GFCI... which recently actually protected me from shock when my 10+ year old fridge developed a ground fault three weeks ago- apparently compressors can develop ground faults when they are near failure. We had a pretty big ordeal replacing it with something brand new from the store... So... a bit disappointing. What I would like to do is install the inverter as the ATS- running only inverter output to the AC panel and the shore inlet to inverter input, so that I can take advantage of ATS/charge, and later have computer control over ATS/inverter modes. It would also render me immune to problems like open neutral at faulty pedestals. Perhaps I could buy a newer revision board with the double pole relay and retrofit it later... hopefully for the time being I don't have to plugin at all =).
  16. Today I was able to replace the output terminals: Getting ready to hook it up! I did acquire some of the "new connectors" currently being used but decided against using them. They are quite a bit smaller, which is nice, but the current terminals are far beefier than anything I've had on a PJ inverter. Finally I built some 4/0 cables to hook it all up: The new 43kWh LFP bank is charged full: 35kWh into the bank, so SoC was around 20% on arrival.
  17. Its not like either of you did it... Both covers were fine, oddly enough. I have to say this is a very nice looking inverter, probably the nicest I've seen thus far! Stainless and shiny! Finished swapping out the batteries today, but ran into trouble with the Daly Smart BMS. I must be missing a temp sensor or something, as there's a related fault code on the BMS preventing charge/discharge, and the one sensor reads -40C. I'm reaching out to Daly about it but I have my doubts as to whether they'll be responsive... So for now I'm going to charge to 54.4V, and keep an eye on the cells manually while I look for a new BMS.
  18. @Sid Genetry Solar @Sean Genetry Solar So, some unfortunate news- seems the front output terminals of my 6kW GS inverter were busted in shipping. Could you send another matching terminal block?
  19. Yesterday we finished base assembly: Now I have some cleanup to do in the electrical cabinet, reorg/moving stuff. That used to be rather neat- incrementalism is to blame. The hanging positive leads go nowhere at the moment- they were to the PJ inverter but will be used now instead to go to the new batteries. Also intend to vacate the lower section today, for the GS inverter: 4 Tesla packs on the right, NAS/landslide of electrical goodies that never belonged there on the left. New home for the GS inverter. Lots of work ahead of me today....
  20. Lower half of the battery bank balance lead bars installed:
  21. So yesterday, the rest of my batteries finally arrived. We started assembly immediately: Custom aluminum bus bars spanning six cells to make both the series and parallel connections in one swoop. Pretty sure these will perform well, if not I'll have some nickel plated copper ones made later on. You can see my masking-tape covered torque wrench at the bottom of the photo above, taking every precaution with assembling these. 48 cells in 16S3P. Behind, inbetween, and in front of these cells I'll be placing some RV heat pads: They have a built-in thermostat, but I'll also add my own themostat to drive both these and a warning light in the cabin. This is my safeguard against freezing. Yesterday I got through about half the cells. BMS is beefy as heck: Using some runs of "PVC wood" to act as barriers between the bus bars (prevent accidents), a platform for a protective cover. I cut channels in these barriers for the BMS balance leads: Today hopefully I can finish up assembling the bank. After that I have to top balance the cells- I'm going to do this by charging the bank at once up to a modest voltage (3.4V per cell, 54.4V bank), then increasing the target bank voltage by .2V until I reach 3.55V for every cell. Any stragglers I'll charge individually with a bench power supply to bring them up. If I have peaking cells, I'll use the bench supply to hit the lowest ones until they are caught up. Top balancing them all in parallel would take far too long from factory SoC with my power supply. After a top balance, I'll remove my Tesla packs and install the shiny new GS inverter, finally!
  22. Performance and parallelization run hand-in-hand if you have lots of computation to do. There's little reason for instance, rendering pixels from a scene cannot be heavily parallelized. Or cryptography, or crunching data in big files. I guess the reasonable steel-man would be to ask what kind of tasks need to be done on general purpose computers that cannot be parallelized well, and why. Is it the task that cannot be parallelized, or is it the technique accomplishing the task that is unsuitable for parallelization? I'd argue the extensive number of special instructions illustrate a weakness in X86, not a strength. Lets say for argument's sake some specific ARM chip manufacturer added transistors to accelerate SSE2 or MMX extensions, its the same thing. The question is, for the mass majority of use cases, do you really need all of those extensions? Or did those extensions come from use cases where X86 fell short? I must be misunderstanding you- Lets say you have some program written in C. Of course you need to compare the native compiled ARM code performance on the ARM chip to native compiled X86 code on X86 variants to get an apples-to-apples comparison, not how well the ARM chip will execute the kind of instructions present in the X86 binaries. Again I must be missing something in your reasoning- Does this apply in reverse? Does X86 emulate / virtualize ARM instruction sets flawlessly at native speeds, cause I never got that memo. Does parallelization come from somewhere other than just more (redundancy) cores? I think the performance comes from IPC. In if one of those 5 billion cycles (5Ghz), you complete one instruction on ARM, but on an X86 chip it takes even just two cycles to complete the same instruction- the ARM chip is going to win by a theoretical 100% margin. I think this is the potential RISC offers- to complete instructions with less cycles than CISC architectures. If you need to do complex work, don't try to make it happen at the instruction set level- make your instruction set as simple, as cheap as possible in terms of cycles consumed and then leave the complex computations to the programmers and their code, where they can parallelize. Its not a functional difference as much as it is a philosophical difference. Disclaimer: I'm no a chip/instruction set expert, I'm a Java programmer, don't even do memory management 😃
  23. XP is where I got off the lolsoft ship. Windows ME seemed like peek awful, then the switch to the NT kernel really stabilized things for a while, only for Vista came along and consumed half the resources on your modern high end gaming rig... which btw can it run Crysis? A lot of people who discover Linux evangelize hard, I used to, but don't anymore. Just tends to annoy others and cause me more work =). Some unfortunate news for @Sean Genetry Solar , like it or not Microsoft is adopting a lot of Linux tech ( https://docs.microsoft.com/en-us/virtualization/windowscontainers/deploy-containers/linux-containers ) even into their core OS ( https://docs.microsoft.com/en-us/windows/wsl/install-win10 ) that professionals and businesses are demanding ( https://www.howtogeek.com/249966/how-to-install-and-use-the-linux-bash-shell-on-windows-10/ ) . Also don't underestimate the effect of Apple switching to the M1 chip or NVIDIA buying ARM, its quite possible that consumer grade computing is on the verge of shifting architectures to ARM wholesale, then Windows will at least temporarily lose its greatest advantage- its immense historical software library. Windows as you know it has only run on a handful of architectures in its history whereas Linux has and continues to run on most prolific architectures, even some obscure ones, it is definitely better poised to compete on ARM than Windows is. Its really kind of silly to compare Windows to Linux... One is a kernel, the other is kernel + userspace + UI + etc... Microsoft could even get out of the game of OSes altogether, opting to ship a "Windows" running the Linux kernel with their own UI on top, and instead contribute to WINE. I'm surprised it doesn't occur to many that WINE could eventually surpass the official Win32/Win64 API implementation entirely, in fact it has in some ways already (like backwards compatibility).
  24. I'm thinking its essential for local network access, without it tokens/authentication is meaningless. TLS or not I wouldn't expose the inverter endpoints directly over the internet. Right, totally onboard with this. I see no need to validate the certificate at all on the server, the clients are going to decide whether or not its valid on their own. and for any self-signed certificate the trust chain is going to be broken by default. The certificate exists to establish a chain of trust as to prevent someone from spoofing your service. genetrysolar.com's certificate chain prevents an attacker from pretending to be genetrysolar.com through the private/public key mechanism and trust chain. In the case of the inverter, most people I'd imagine would be using self-signed certificates where there is no root trust authority, with certificate expirys set well into the future as to prevent headache.
  25. The forums' first flame war has erupted! (just kidding) This is saying "its okay if the inverter is insecure over LAN because other stuff on your network is also likely to be insecure", which is precisely NOT what I was saying. Additionally being able to discover my printer and print to it without my authorization is a different animal than being able to change inverter states. I was saying that even if link-layer security fails (wifi password, etc), people still have transport level security that prevents traffic content from being sniffed. Exposing the inverter over the internet is also a bad idea even with TLS, I wouldn't do it. Now, it could be that the limitations of the WiFi board are what they are, encryption for this is too much, etc. That's certainly plausible. It may be that dealing with certificates/etc just makes implementing this impractical or the effort/reward isn't that, or a variety of other reasons this cannot be done. I'm just suggesting ways it could in theory be done. Its not the end of the world in this case- without TLS I'd put the inverter on an adhoc network with a device (Raspberry Pi) that acts as a reverse proxy and adds TLS to it all.
  • Create New...