PLEASE NOTE: If you had an account with the previous forum, it has been ported to the new Genetry website!
You will need to reset the password to access the new forum. Click Log In → Forgot Password → enter your username or forum email address → click Email Reset Link.
As with any new design or idea, it's usually an extremely good idea to have a "prototype" stage, before rushing a design into production. I've sorta skipped that in the past on some designs--sometimes with success, sometimes without.
At any rate, while the Rev. C prototype board DOES work for the core functionality of driving FETs for an inverter, I've discovered quite a few mistakes with the design that will need tweaked--and so many in fact, that I'm going to have to do another prototype run just to make sure I get all the issues worked out BEFORE sending it to the manufacturer and asking for 300pcs to be manufactured...
Issues I've discovered include:
- reversed polarity on BOTH AC voltage sensors (input and output)
- completely unusable scale on BOTH external AC current sensors (input and output again)--got my math wrong, and it's 4x what it needs to be
- connected BOTH new AC current sensors to non A/D pins on the CPU, rendering them basically useless
- miscalculated the feedback circuitry for the new LCD power sense chip (using a Chinese power monitor chip now, as the CS5463 is nearly 7 times the cost...and almost impossible to find)
- need to swap 2 A/D inputs on the CPU to allow use of an internal comparator that can instantaneously shut down the FETs if the battery current exceeds a set threshold (12kw safety protection)--and I'm NOT very good at rewiring 0.8mm spaced pins!
- used the ULN2003 for driving the relays, only to discover that it's only good for 0.5A--and the new relays pull 650mA and 1.2A. The latter also overloads the on-board power supply--so I need to switch to a push-pull relay driver IC (L298D being considered right now), and drive the relays in a push-pull "single coil" manner instead of a "double coil" manner (which also reduces their power consumption by 1/2)
On the bright side, I did find that while one of the relays is really tall, it will in fact fit in GS 6kw chassis without hitting the lid (or the LCD board). AND...if I make a slightly modified LF Driver board (probably will have the PCB manufactured bright red so it's obvious!), the Rev. C design will be compatible with the existing GS mainboards/MOS boards. Might need a small resistor bridge board for backwards-compatibility with the current-design WiFi (LCD) boards, and if all goes well then, we should be able to offer Rev. C upgrades if perchance someone feels that they need it.
(For GS 6kw inverters, there's next to no reason for a Rev. C unless you need the upgraded 2-pole input relays with auto voltage switching.)
Did you order them yet?
I wish. Been frankly unexpectedly busy on more personal matters this past week...and I haven't gotten them ordered. They're on the goal/plan for today.
The order of Rev. C board relays shipped yesterday.
Mine will likely be coming back for that board and a status check of things inside. And a new set of fan grills. Mine are already starting to rust which also happened early on with the PJ/UPower ones. You and Sean might see some other things inside that the humidity will have affected. 97% humidity right now which may end up doing things to the board or components. ( Told you it was Damp down here ) The MingHe boost controllers have about a 6 month life because of it so my unit is a good test for humidity affects.
Mine will likely be coming back for that board and a status check of things inside. And a new set of fan grills. Mine are already starting to rust which also happened early on with the PJ/UPower ones. You and Sean might see some other things inside that the humidity will have affected. 97% humidity right now which may end up doing things to the board or components. ( Told you it was Damp down here ) The MingHe boost controllers have about a 6 month life because of it so my unit is a good test for humidity affects.
Oh my, that's a bit more of a test than I expected ;-).
Having to find a different PCB supplier, as JLCPCB basically told me that they don't care to keep parts in stock anymore. I really don't care for SMT soldering that much, so I'll have to use someone else. Considering a different PCB supplier (costs more, of course!) that offers conformal coating. That might be highly important for reliability in a humid environment.
Uhm dang waterman, might want to consider, enclosing those inverters in a grow tent with several dehumidity in it to keep the humidit ydown.
Uhm dang waterman, might want to consider, enclosing those inverters in a grow tent with several dehumidity in it to keep the humidit ydown.
Then we would get hit with the other problem. Daytime temps are in the upper 90s. Often we get hit with a double whammy of temps and humidity. At 2030 tonight it is 81 with a 75% humidity. We run a dehumidifier in the house along with the A/C. Up to 3 gallons a day come out of the dehumidifier. The A/C unit often has a stream of water instead of dripping.
Finally got the Rev. C control boards / WiFi boards ordered. A LOT later than I hoped, but hey, they're ordered. Now hoping I didn't make any stupid mistakes on the boards...