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.
A quick check with my phone of the relays on the PREVIOUS board design we had the factory make (Rev. B) turned up an absolute maximum reading of 50uT. My phone maxes out at 200uT with the Rev. C board relays.
According to the above link, the maximum allowable magnetic field from a package is 200nT (or 0.2uT) at 7 feet.
Pretty sure that's the problem....
Wait... what did your phone say at 7 feet?
Magnetic field strength will decay at a quadratic rate relative to distance. Wouldn't it be much lower than necessary by the time you reach 7 feet?
Wait... what did your phone say at 7 feet?
Well, it's almost impossible to determine the maximum 0.2uT allowed magnetic strength with a consumer-grade device when according to Wikipedia, the earth's magnetic field strength ranges from 25-65uT (which you can easily see on my phone screenshot above, with the "ambient" under 50uT).
However, someone might be able to mathematically calculate the decay of 200uT from point-blank to 7 feet?
Define Point-Blank...
If this is like gravity, it would decay at something like: f/r²
Given point-blank, r = 7/pb
At 1ft, 200/7² = ~ 4uT
At 1/2ft: 200/14² = ~ 1uT
At 1/4ft: 200/28² = ~0.25uT
At 1/8ft: 200/56² = ~0.06uT
Indeed the formula is the same inverse-square law as used by gravity.
https://www.instructables.com/Evaluate-magnetic-field-variation-with-distance/
200uT @ 1" would be 200uT/(7*12)-1)² = ~0.03uT @ 7'
That is, of course, considerably lower than the 0.2uT threshold. Huh, maybe they didn't scan it right or your tool is just not accurate enough. 😕 An order of magnitude is an awful lot of error, though...
ONE of the 2 packages arrived...it's the 6kw case prototype. This means that I got a good look at the battery terminals. I'm not 100% sure, but wouldn't be surprised if they're at least mostly copper. Took a file to the edge of one of them, and it's got the same reddish color inside as it does on the outside. Compare to an (obviously) brass GS12 terminal and see the diff....
Heck yeah that looks like solid copper to me or at least a much better alloy than the 12K ones. That is good news man.
Guess what: this is the VERY first time that we have had the factory ship a Rev. C control board--which has 2 large latching relays. These relays have very strong internal magnets to "hold" the state with no power.
That is a good guess, I had a similar issue with a Lift system control board they also have magnetic latching relays. From how it was explained to me is they have a wand that they run over the packages if it goes off they pull no fancy meter or anything like that. Maybe others do though, but doubt that they walk around and measure things.
September 4th:
- The re-shipped 12kw prototype unit has reached Los Angeles...it's finally in the States. ETA is September 9th.
- I've done a bit of testing on some revised FET boards with an active Miller clamp circuit...and let's just say, the initial tests are VERY promising! If these FET boards do well at all the tests I can throw at them, they're going in ALL the production GS inverters from this point forwards (i.e. both 6kw and 12kw). This includes all of the preorder inverters.
Sid, when you get a chance can you post the rough measurements of the inverter cases. I am working on the spot for my 6K, but curious about both of the new ones.
Sid, when you get a chance can you post the rough measurements of the inverter cases. I am working on the spot for my 6K, but curious about both of the new ones.
New GS6 inverter chassis:
- height (including 25mm foot): 185mm
- width (NOT including 25mm foot on both sides): 290mm
- length (NOT including chassis connectors): 420mm
- don't have a weight yet, but the shell (without any boards, transformer, or fans...came in at 18lbs shipped!)
First production GS12 inverter chassis:
- height (including 25mm foot): 205mm
- width (NOT including 25mm foot on both sides): 385mm
- length (NOT including chassis connectors): 500mm
- don't have a weight on this one yet...but as soon as the prototype arrives, I'll know pretty close!
You might note that the case sizes seem not that much different. But from a volumetric standpoint, the GS6 case has 19,488 cubic centimeters of chassis volume, while the GS12 case is 34,650 cubic centimeters. (A significantly higher transformer turns/volt ratio in the GS12 makes the size not have to be exactly double.)
September 6, 2022:
- Actually received the 12kw "test manufacture" unit. They actually did quite well...just a few things to address, and a lot of things to test/verify, plus a few things to adjust.
- If all goes well, I will be able to give the factory an "all clear" to start ordering parts next week for the production run. About time!
- Shipped weight of the 12kw measures at 111lbs. I would expect a shipping weight closer to 115lbs (as units coming from the factory do NOT have a WiFi board in them, and the ferrites for this inverter went in 6kw "shell" box). Still, when ALL the competition has to ship freight due to being >150lbs, I think we're doing pretty good!
September 7, 2022:
- Had a good look over the GS 12kw test unit. Turns out there were quite a few more "wrong screw" issues--but nothing structural/functional as much as "not what I specified."
- Planning to ohm out the transformer tomorrow and see how much lower the losses are with the thicker wire.
- Got the updated "active clamp" FET boards installed in the GS 12kw. Pending testing starting hopefully tomorrow.
- Discovered that the "must be made of gold" mainboard caps are actually better than the EOL caps I bought from Mouser! They have a notably lower ESR, though I cannot find a datasheet to get the ripple current rating...
September 8, 2022: It's alive!
- Finished measurements, tests, tweaks and adjustments.
- Reassembly pretty well complete. Fired it up...a cool no-load of 44W. Please note that this is without the 4 cooling fans installed; they do add a little quiescent. Still, expected no-load is <50W. Compare to "the big names" at 150W and up for no-load draw!
(yeah, it's a tad blurry, but...!)
I am hoping to get all the design files updated based on my findings, and out to the factory by the end of this week. Monday (their time) they get the green light to do the production run.
FWIW going to the thicker transformer wire reduced the resistance roughly 17% on both windings. Losses similarly decreased by 17%--but considering how small the losses are compared to the total power rating, the total "loss-calculated" efficiency went from 94.76% to 95.58%. Note that this efficiency does NOT indicate total system efficiency...but rather simply calculating power transfer through the transformer, less the measured DC resistance losses. Actual efficiency will be lower due to other system losses.
Curious about those current sensors, it looks like there are multiple wires running through both of them, but that could be an optical illusion.