Grid-tie Master Mode

32 minutes ago, Cali Carlos said:

<a contenteditable="false" data-ipshover="" data-ipshover-target="/profile/2-sid-genetry-solar/?do=hovercard" data-mentionid="2" href="/profile/2-sid-genetry-solar/" rel="">@Sid Genetry Solar I was wonder how you felt the gen2 12v 6k inverter would do on a 3k grid tied system without linear throttle? We talked a bit back and fourth over a year ago and I was wondering if anything new has come up with it? My grid tied inverters will shut down for 5 minutes if the frequency reached 60.5 Hz for .16 seconds which I'm fine with. Due to inherent inefficiencies my system maxes out at around 2.5kw and I have roughly 300w of always on loads so the inverter would have to be able to charge 2.2kw at 12v on a worst case scenario. I feel the big issue would be the high charge current. Do you feel that's at all feasible indefinitely or for X amount of minutes/seconds? It was my understanding that if the inverter got to hot due to the charge current that it would shut down the grid tied inverters before it shut itself off?

(going back to reread the conversation then...)

So some of the same limitations DO still apply.  However, due to the significant monitoring improvements on the Rev. C board...there's also some significant improvement as well.

In the 6kw inverter, the FET max is conservatively estimated at ~600A.  That's pretty close to 6kw with no surge--but hey, that's 600 amps at 12v!  Wire losses will likely be a pretty significant limitation here.

As far as grid-tie charging, I do not see any issue with even 3kw of charging @ 12v...that's ~300A, well below the conservative max.  The important factor to keep in mind with grid-tie charging is that the FET body diodes are NOT utilized--and as such, the heat/losses will be very low. 

(Standard AC Input charge--i.e. from generator--does use the body diodes on the FETs, which can easily drop 1.2v across the entire bridge * 300A = 360W of heat.

As of right now, if the inverter gets too hot, it will shut the entire system down.  HOWEVER, the GTM function is in extreme infancy right now (mostly due to not having a test bed to experiment with) and if grid-tie charge proves problematic with charging, it would be very easy to add a freq. shift shutdown if backfeed charge is active and temps are getting high.  Firmware adjustment, not a hardware adjustment--so very easy to implement in the field.

One thing that HAS changed for the better is that now with Rev. C boards, the inverter is fully aware of total power flow on output AND power input--not just the amount of power, but also the direction of power flow.  This means that it can detect "backfeed power" into the generator--and while it is completely unable to adjust the frequency in pass-thru mode, it would be feasibly possible to have the inverter shut down.  Or...I mean, the sky's kinda the limit here...it could be configured to automatically shut the generator off...which will result in a drop back to battery mode and then the automatic grid-tie shut down.

And also, as you may have noticed, the AC input on the GS inverters now can auto-switch between 120v or 240v input, with separately defined amperage limits for both states.  Handy for campers!

Does that mean that the 12v version doesn't have much of a surge capacity above 6kw?

So basically if it doesn't work out of the box you feel that with some troubleshooting and diagnostics you'd be able to get it to work?

13 minutes ago, Cali Carlos said:

Does that mean that the 12v version doesn't have much of a surge capacity above 6kw?

Pretty much.  I don't think this is going to be much of an issue on the inverter's side--because any sort of surge capacity calls for borderline extreme/insane amperages.  The likelihood of batteries and cables being able to supply said currents without significant voltage drop is extremely low. 

For example, if the entire power path (including the batteries' internal resistance), battery cables, inverter wiring, FETs, transformer primary, etc. comes out to a relatively miniscule 0.012 ohms (that's just over 1/100th of 1 ohm), Ohm's law shows that 1,000A across 0.012 ohms = 12v.  This means that the DC losses alone will consume all of the available voltage--leaving no power to do any work (like power a load).  That's a mathematical metaphor, obviously--power doesn't quite work that way, and the system simply won't be able to reach 1,000A if the total DC resistance is 0.012 ohms.

Conversely, if you're running at 600A (the conservative maximum of the FETs), if the ENTIRE system resistance is 0.005 ohms (5 milliohms), you have a 3v loss--so now instead of getting 12v as necessary, the transformer will only be seeing 9v.  This is the issue with low voltage systems...you run out of headroom VERY quickly.  A 3v loss at 55v simply brings you to 52v--and you could literally care less.  But a 3v loss at 12v is literally 1/4 of the total voltage.

25 minutes ago, Cali Carlos said:

So basically if it doesn't work out of the box you feel that with some troubleshooting and diagnostics you'd be able to get it to work?

I can kinda guarantee you it likely won't work "out of the box" like you have in mind (depending on how much code work I get done between now and then!)--but I can similarly guarantee you that with some back-and-forth (diagnostics, debugging, tweaking, adding code, etc.) we will be able to get it to work as you have in mind.  The hardware is completely capable...it's just going to take some diagnotics/testing/tweaking/adjusting via firmware updates and diagnostic screens to get what you want.  Rev. C's comprehensive power monitoring abilities make it completely possible.

Check out this screenshot of the local server for the upcoming 1.2r0 firmware.  Note that the Output AC Amps, Input AC Amps and Transformer Amps all can read both positive or negative?  This makes it very easy to determine the direction of power flow--if the grid-tie inverters are backcharging the batteries, the Transformer Amps will go negative.  (So will Output Amps.)  And if the inverter is in pass-thru, you can similarly see the Input Amps go negative if the grid-ties are backfeeding the generator.  Very easy for the inverter to react to those cases.

<img class="ipsImage ipsImage_thumbnailed" data-fileid="1482" data-ratio="82.00" style="width:400px;height:auto;" width="906" alt="image.png.000276a79d65c25eb2518749cc23bc83.png" data-src="/monthly_2022_07/image.png.000276a79d65c25eb2518749cc23bc83.png" src="/applications/core/interface/js/spacer.png" />

And if the inverter is in pass-thru, you can similarly see the Input Amps go negative if the grid-ties are backfeeding the generator.  Very easy for the inverter to react to those cases.                     This mean the  relay will shut down the inverter  if someone  backfeed grid power  to the  inverter output by mistake   when the inverter is running   .  The  relay need to   shut down the inverter in  1/60 second  and save the FETs .    IF  that is possible then  there is no way to destroy the inverter  .     

11 hours ago, Sid Genetry Solar said:

-but I can similarly guarantee you that with some back-and-forth (diagnostics, debugging, tweaking, adding code, etc.) we will be able to get it to work as you have in mind.  The hardware is completely capable...it's just going to take some diagnotics/testing/tweaking/adjusting via firmware updates and diagnostic screens to get what you want.  Rev. C's comprehensive power monitoring abilities make it completely possible

And I'd definitely help in whatever way I can.

Now just to decide what dc voltage I want. I was leaning towards 12v because I'd be able to "jump" it from my truck incase I needed to supplement power (unorthodox inefficient and definitely not green I know) but it would only be for the extremely rare case that we had a long term power outage.

What would it take to run my ac unit pictured below? Higher dc voltage? A soft start? Or the 12k? I don't remember the size of Sean's smaller unit.

<a class="ipsAttachLink ipsAttachLink_image" href="/monthly_2022_07/20220718_184514.jpg.265a0ee4f5118dfe793d162e931d844b.jpg" data-fileid="1485" data-fileext="jpg" rel=""><img class="ipsImage ipsImage_thumbnailed" data-fileid="1485" data-ratio="133.45" width="562" alt="20220718_184514.thumb.jpg.973c673d033612a9b39338b5e541e5b1.jpg" data-src="/monthly_2022_07/20220718_184514.thumb.jpg.973c673d033612a9b39338b5e541e5b1.jpg" src="/applications/core/interface/js/spacer.png" />

11 hours ago, Sid Genetry Solar said:

And if the inverter is in pass-thru, you can similarly see the Input Amps go negative if the grid-ties are backfeeding the generator.  Very easy for the inverter to react to those cases.

Can the batteries be backfed in pass-thru?. If it's just a basic manual generator without auto on/off would there be anything the inverter could do to stop the backfeed into the generator? For example using the ATS feature to disconnect from the gen, raise the hz to shutdown the grid ties and then connect back to the gen?

1 hour ago, Cali Carlos said:

What would it take to run my ac unit pictured below? Higher dc voltage? A soft start? Or the 12k? I don't remember the size of Sean's smaller unit.

If Sean's tests are any benchmark, a 6kw on 48v should be able to start that A/C (yours has a spec of 68 LRA; Sean's small A/C has an LRA of 77A).  We have not tested this on a lower voltage (i.e. 24v), though mathematically a 24v unit would likely be able to start it, especially if a soft start reduces the amperage.

Mathematically an LRA of 68A * 240v = 16,300W at startup / 48v = ~340A.  At 24v, though, 16,300W / 24v = 680A--and the FETs are going to limit ~600A for their safety.  Note however that this doesn't mean it won't work, just that you're on the mathematical redline.

Note that the FETs' 600A limit does not change based on the battery voltage--so a 48v unit will have double the surge of a 24v unit, and four times the surge of a 12v unit.  Also note that there are not necessarily "better FETs" that can be utilized for 12v operation: the real limitation is melting the leads off of the TO-220 package.

Sean's bigger A/C has an LRA of 109, and a 6kw cannot start it--it just hums angrily and won't turn over.  (Likely the DC losses of trying to run over 3x the rated load out of the transformer result in just simply not enough power left over.)

Conversely, if you had a variable-speed compressor OR a mini-split (with an "inverter" compressor), there would be no issue running any of this on even a 12v system.  Inverter compressors have no startup surge.

2 hours ago, Cali Carlos said:

Can the batteries be backfed in pass-thru?. If it's just a basic manual generator without auto on/off would there be anything the inverter could do to stop the backfeed into the generator? For example using the ATS feature to disconnect from the gen, raise the hz to shutdown the grid ties and then connect back to the gen?

Batteries can be charged in "AC Input" mode from the genny, yes--but that is different from backfeed charging in inverter mode.  The inverter can technically pull up to the full 6kw from the genny as charge--so I don't think that'll be a problem (unless internal heat becomes an issue, in which case the inverter will linearly throttle down charge as necessary).

Depending on the type of genset you have, a backfeed might not be too much of a problem.  I can almost envision a "crate genny" with no electronics on the output...getting backfed, which tries to actually power the engine through the generator, and you end up not burning any gas at all 😉.  However, it almost certainly doesn't work that way...but the concept is pretty funny.  If it did, however, it'd likely end up pushing the frequency up and the grid-ties would pop themselves off.

Up to testing to find out what needs to be done.  I don't think you'd quite appreciate an ATS disconnect every 5 minutes to pop the rogue GTs off...but, it would be technically possible.

Trying to protect from an AC backfeed while it's running is almost downright impossible.  1/60th of a second is WAY too slow to prevent FETs from exploding in that case. 

Output short-circuit protection is one challenge, comparable to running a vehicle head-on into a concrete barrier at 60MPH.

AC backfeed is a completely different challenge--equivalent to running a vehicle at 60MPH head-on into a concrete barrier which is simultaneously moving towards the vehicle at 60MPH.  It's potentially at least twice as bad as a short-circuit, depending on how the AC waves happen to either "match" or "collide."

Output short-circuit protection is one challenge, comparable to running a vehicle head-on into a concrete barrier at 60MPH.

I  short circuit my  PJ 15 kw  2 years ago  and Sean  short circuit  his inverter also  and it still works .  I like to see a test of backfeed to inverter output while running  and still survive .  I  see the video of  the GS 6kw  returned  to replaced all the parts by someone  who admit he backfeed .       What if he did not admit his mistake  then the  GS 12kw need a  data  recorder  to record the last  minute  before all the FETs blow up .     

2 hours ago, Sid Genetry Solar said:

Batteries can be charged in "AC Input" mode from the genny, yes--but that is different from backfeed charging in inverter mode.  The inverter can technically pull up to the full 6kw from the genny as charge

Can a smaller generator be used on the input side to strictly charge the batteries while the inverter pulls power from the batteries, basically using the batteries as capacitors for the larger shorter loads while the generator keeps the batteries charged? If so can it be used in conjunction with GTM or strictly inverter mode?

2 hours ago, Sid Genetry Solar said:

Up to testing to find out what needs to be done.  I don't think you'd quite appreciate an ATS disconnect every 5 minutes to pop the rogue GTs off...but, it would be technically possible.

Definitely not something to do over and over again on purpose, just more of a failsafe

43 minutes ago, Cali Carlos said:

Can a smaller generator be used on the input side to strictly charge the batteries while the inverter pulls power from the batteries, basically using the batteries as capacitors for the larger shorter loads while the generator keeps the batteries charged? If so can it be used in conjunction with GTM or strictly inverter mode?

Only if it's a DC generator connected directly to the batteries.

If you're talking of an AC generator connected to the inverter, then no, not possible.  Power can only flow one direction, it cannot flow both ways at the same time.

You can either charge the batteries, or discharge them.  It's not possible to do both simultaneously.

However, I think I know what you're really asking about, and I do have plans for a "hybrid sharing" function.  In this case, you can set the AC Input limit to, say, 10A (or whatever the max of your genny is).  And then half an hour later you turn on the 15A microwave.  The inverter would quickly throttle from AC charging (if enabled) negative into "power boost", and pull the extra 5A of power from the batteries...regulate the input at 10A...and then when the microwave turns off, quickly throttle back into charge again.

That function is not implemented yet, but fully possible.

EDIT: this function would not support frequency shift.  Anything using AC input cannot frequency shift--because then the inverter is not responsible for the AC line frequency.

9 hours ago, Sid Genetry Solar said:

However, I think I know what you're really asking about, and I do have plans for a "hybrid sharing" function.  In this case, you can set the AC Input limit to, say, 10A (or whatever the max of your genny is).  And then half an hour later you turn on the 15A microwave.  The inverter would quickly throttle from AC charging (if enabled) negative into "power boost", and pull the extra 5A of power from the batteries...regulate the input at 10A...and then when the microwave turns off, quickly throttle back into charge again.

Yes that's exactly what I meant, sorry. So set the max input of say 1200 watts,  if there's a 700 watt load then 500 watts goes into the batteries and if there's a 3000 watt load then 1800 comes out of the batteries. All roughly of course, not figuring losses. Am I understanding correctly?

How would the inverter feel about a modified sine wave input? And if it's ok with it would it output the same modified sine wave or clean it up?

That is the goal.  (It'd be based on amps, not watts.)  I have yet to implement that function, but it's on the list.

I don't know...that would be interesting actually.

It will output the same.  There is no hardware ability to "filter" AC Input power.

Trying to protect from an AC backfeed while it's running is almost downright impossible.  1/60th of a second is WAY too slow to prevent FETs from exploding in that case. 

I  watch  the 5  hours  video of the 12kw GS  comments  and  Sean  say  it shut down  in 1/60 seconds  wnen  backfeed is detected  on the  inverter output  ( a reverse AC OUT and shut down )   so it  did the impossible  and save the FETs .     

<a href="/monthly_2022_07/1721438980_Screenshot(616922).png.7e2c34c88261456a40220d610b06426a.png" class="ipsAttachLink ipsAttachLink_image"><img data-fileid="1486" src="//forums.genetrysolar.com/applications/core/interface/js/spacer.png" data-src="/monthly_2022_07/78663418_Screenshot(616922).thumb.png.6fa4aebff3e96bf342162ebb8468d55e.png" data-ratio="56.2" width="1000" class="ipsImage ipsImage_thumbnailed" alt="Screenshot (616922).png">

Nope.  That's not what Sean was referring to...and he never said 1/60th of a second.

The issue that Sean was referring to there is in an ATS transition when the relay cuts out.  Sometimes due to loading on the inverter, the voltage feedback briefly registers negative (i.e. out-of-phase) voltage on the line for a cycle or 2.  Due to a zero-tolerance firmware, that causes the inverter to throw an "Xformer Polarity" error and shut down.  There's no actual problem, so I've tweaked the firmware to be a bit more tolerant.