powerjack 20,000 watt split-phase PSW ASL 11.0 48-volt parts machine

me either.. I would not rewind an ASL9 transformer....I did see Sean was selling it on his GS web site for $349 for 48-volt version and $499 for 24-volt ASL9 transformer plus shipping w/o warranty and said they are prized and rare??? wonder what the difference is????

The ASL9.0  is a lot taller  and the  the  secondary  winding  is a bundle of wires crimp to make  it a larger  diameter  then the  secondary  diameter of the AS9 transformer  .  The  secondary winding  is very difficult to rewind  so the ASL9.0  is worth a  lot more .    

19 hours ago, Sid Genetry Solar said:

If this is the Jack Daniel video you're referring to, there's several major issues:

NO it is not the same video i was referring to as it was only 30 day old and talking of the new improved 2022 version...evidently they have made some changes...

the youtube video of John daniel i  was watching was comparing the old to the new cnswipower 3000 watt PSW HF inverter upgrades. he was talking about the 2022 version with upgrades vs the old version and i do not see a load test...

but as you have already said you had bad experience with the 2017 model you bought, and do not like them.

PowerJack 20000 watt LF PSW SP parts machine Time to do the green wire repair today!!!

I may use the head lamp to get a bit more light on the project.

Black Friday 2022 is here and the deals are ho hum...think I will just stay home and not waste the time and gas...🤣

43 minutes ago, pilgrimvalley said:

NO it is not the same video i was referring to as it was only 30 day old and talking of the new improved 2022 version...evidently they have made some changes...

the youtube video of John daniel i  was watching was comparing the old to the new cnswipower 3000 watt PSW HF inverter upgrades. he was talking about the 2022 version with upgrades vs the old version and i do not see a load test...

I referenced the video you linked in the bottom of my previous post here:

20 hours ago, Sid Genetry Solar said:

Jack Daniel's new video from a month ago runs to a MAXIMUM load of 1,440W.  Again, this is NOT a proper load test of the inverter! 

44 minutes ago, pilgrimvalley said:

but as you have already said you had bad experience with the 2017 model you bought, and do not like them.

And I don't see any hardware changes in John Daniel's videos that would fix this issue (quite the opposite with the visible cost-cutting measures)--nor do I see any tests proving that the problem(s) have been fixed.

Yes, they're very cheap.  But again, you get what you pay for.

yes, >>>>> PowerJack is a stab in the dark in what they send out, and when it may actually be delivered....sometimes they send only the fraudulent piece of paper , other times they send other inverters of constantly changing specifications...

if you can show details of how you test it and pictures...that would be very helpful....   i am more into carpentry and still learning more about electronics,,

The  picture show how I test the mosboard and  FETs  .    The mulimeter is in Diode mode .   The first leg is always 1 .   The third leg is  between  450 to 650 .     The  mosboard in my  9000 kw have all  green LED light on .   The mainboard is good and worth  200 dollars .    I will test the  control board  later .    

Good common sense that's not often talked about.  Inverter and battery manufacturers just want to push more products.  Seems silly that people would have big A/C off-grid without soft start which is likely more economical overall than having a bigger inverter/battery system to start the A/C.

On 11/23/2022 at 9:51 AM, Sid Genetry Solar said:

 

Yes, they will do the full 3,000W output, with 2 caveats:

• MUST be a purely resistive load.  These inverters are absolutely horrible at handling any sort of inductive load.  And I do mean terrible.  A 120v T20 stapler I have won't even break a staple off the pack.  (Compare to a Meanwell TS-3000 HF inverter that I had beforehand, which did run the stapler OK, if slightly weaker than the grid.)  If you're running a fridge/freezer (which btw is the biggest compressor the Swipower can start), ANY inductive load will cause a dropout big enough to stall the fridge/freezer compressor.  Which then will sit locked up until the compressor's thermal protector trips out.  And (from experience) those thermal protectors are not designed to be regularly used--they WILL fail.
• The SwiPower will only run 3kw for about 20 seconds before overloading and shutting off.  Electrically it's fully capable of running 3kw...just the overload circuit is too trigger-happy.  Max continuous I was able to get was about 2,700W--which fails their advertising of, "It will do the rated power, unlike everyone else..."

Frankly, even that's being generous.  What was the output voltage when you ran 3,000W resistive load?  Most of the HF inverters are way overrated.  Looking the design of the unit it's true max continuous load capacity is at most 2,500W with nominally safe/acceptable output voltage.  Probably good to keep load under 2,000W even for resistive loads to maximize life span.  The reality is that HF units need to be significantly derated (50%+?) to meaningful comparison with LF units.  It would be nice if the industry could develop a consensus standard on inverter testing like max continuous load with output voltage sag < 5V or max surge load with output voltage sag < 20%.

On 11/23/2022 at 9:51 AM, Sid Genetry Solar said:

 

These units use the EG8010 chipset for the output SPWM driver.

 

The EG8010 is pretty good for what it is but has the inherent limitations especially for fast load transients due to peak voltage based control.  How is Genetry inverters different/better in terms of output voltage level and waveform control for fast load transients?

40 minutes ago, JIT said:

Frankly, even that's being generous.  What was the output voltage when you ran 3,000W resistive load?

It maintained voltage just fine--until the "overload trip" engaged.  That's why I say that it was technically capable of doing the rated power.

The reason that HF inverters fail surge loading so badly...is because the boost converter section is generally only rated for about the maximum load in the first place.  This means that the only "surge" capability you're going to get, is from the DC bus filter caps--which let me tell you, isn't worth anything!

Adding to that, the SwiPower inverters' output IGBT bridge current limit (yes they do have one, as part of the basic EG8010 circuit) cuts in far too soon, which results in the sliced-up AC wave visible in my video above.  If this was tamed down a bit, it'd be a FAR more powerful inverter.  Not sure about reliable, just more powerful 🤪.

case in point, in my hacking on one of the SwiPower inverters, I disabled said IGBT current limit.  Unfortunately, this was after I'd damaged the boost converter voltage regulation feedback...so the boost converters didn't throttle up.  But when tested with the T20 stapler, the stapler had quite a kick--as much as the DC bus caps had, anyway.  Significantly more powerful than in the stock config.

44 minutes ago, JIT said:

The EG8010 is pretty good for what it is but has the inherent limitations especially for fast load transients due to peak voltage based control.  How is Genetry inverters different/better in terms of output voltage level and waveform control for fast load transients?

So it is hard to do a direct comparison between an HF inverter with a wimpy boost converter, and an LF inverter whose throttle regulation has to change far more due to battery voltage than for actual load!

I can't say the GS inverters are as fast as they potentially COULD be--but the risk of speeding up/tightening up the regulator response ("single cycle response would be great!") is that the regulator tends to devolve into a wild oscillation of overcorrections.  Which we HAVE seen on our bench, as well as had customer complaints about ("It won't start a window A/C, but buzzes really loudly!  I was afraid it would blow up!")  A firmware upgrade to a less-sharp output regulator algorithm solves that issue.

On 11/26/2022 at 1:52 PM, Sid Genetry Solar said:

The reason that HF inverters fail surge loading so badly...is because the boost converter section is generally only rated for about the maximum load in the first place.  This means that the only "surge" capability you're going to get, is from the DC bus filter caps--which let me tell you, isn't worth anything!

It's often worse than just underrated design.  Don't know about SWI but the DC-DC stage in a cheap HF inverter is typically a push-pull topology boost converter with parallel MOSFET's and transformers that are under-spec'd with little or no snubbers and driven continuously in avalanche mode that degrade the MOSFET's every cycle until one day they just randomly blow up from the accumulated damages.

On 11/26/2022 at 1:52 PM, Sid Genetry Solar said:

Adding to that, the SwiPower inverters' output IGBT bridge current limit (yes they do have one, as part of the basic EG8010 circuit) cuts in far too soon, which results in the sliced-up AC wave visible in my video above.  If this was tamed down a bit, it'd be a FAR more powerful inverter.  Not sure about reliable, just more powerful 🤪.

That's known issue with EG8010 even for LF inverters using EG8010.  I believe the Backshed inverters using EG8010 often disconnect the overcurrent feedback to avoid nuisance trips.

On 11/26/2022 at 1:52 PM, Sid Genetry Solar said:

So it is hard to do a direct comparison between an HF inverter with a wimpy boost converter, and an LF inverter whose throttle regulation has to change far more due to battery voltage than for actual load!

Well ... battery voltage and load are often tightly linked.

On 11/26/2022 at 1:52 PM, Sid Genetry Solar said:

I can't say the GS inverters are as fast as they potentially COULD be--but the risk of speeding up/tightening up the regulator response ("single cycle response would be great!") is that the regulator tends to devolve into a wild oscillation of overcorrections.  Which we HAVE seen on our bench, as well as had customer complaints about ("It won't start a window A/C, but buzzes really loudly!  I was afraid it would blow up!")  A firmware upgrade to a less-sharp output regulator algorithm solves that issue.

Ideally the control loop update frequency should be at least an order of magnitude higher than the fastest transient frequency to be compensated for to minimize risk of instability.

16 minutes ago, JIT said:

Ideally the control loop update frequency should be at least an order of magnitude higher than the fastest transient frequency to be compensated for to minimize risk of instability.

I don't see any way for the control loop to run any faster than 120Hz (i.e. once per half wave), as the entire wave has to be measured before a decision can be made.  And even that runs risk of going into oscillation with high/low wave (ESPECIALLY if you've got a heat gun with a "medium heat" that just puts a diode in the circuit, clipping off half of the waveform).

worth noting that the PJ inverter regulation runs a perpetual +1/-1, so one wave is always a tad higher than the other--and one side of the H-bridge runs just a little hotter than the other 😉.

On 11/25/2022 at 6:59 AM, Sid Genetry Solar said:

Yes, they're very cheap.  But again, you get what you pay for.

yes, >>>>> PowerJack is a stab in the dark in what they send out, and when it may actually be delivered....sometimes they send only the fraudulent piece of paper , other times they send other inverters of constantly changing specifications...

the HF 3000 watt 6000watt peak cnswipower inverter is to be used like most inverters at less than their full rated power..... for my use that would be lights and some outlets of occasional use and not try to run at 3000 watts anyway for that subpanel....John Daniel indicated he has used and abused his 10 plus year old cnswipower 3000 watt inverter and it is still working, he reviews a newer 2022 version in his YouTube video but does not show an overload test....that is for the people who don't care if they destroy their inverters....for me >>> I just want it to work or utilize as a standby unit for when I want it to work and will not run it to full rated capacity....thanks for that warning

in my off-grid solar builds >>> the inverter is minor in price compared to the LiFePo4 battery and the solar panel investments..... unless one opts for the high priced inverters.... for an inefficient appliance band-aid... vs energy conservation >>> energy conservation should come 1st ( my way of thinking )....

I have obtained all inverters at significantly lower prices and they seem to be working without issues....

but then again I do not have a well pump or an in-efficient hard to start central air conditioning unit.....in my off-grid solar build....nor do I have plans to get these in-efficient units...quite the contrary....those are luxury for those who hire someone else to do it all and are not at all concerned about their budget or appliance efficiency... although the well pump may be a critical item >>> a solar powered well pump would be easier to power so update the well pump to a more efficient pump 1st if you need it...

so, the $470.23 20,000 watt PSW LF SP PowerJack parts machine still may not be efficient but it is less than 1/3 the price of its competitor even if it will only do 7000 watts continuously.... definitely get what you pay for....every PowerJack comes in a different version (either upgrades or downgrades). have not tested all of the parts machine just yet >>> but it will be a back-up inverter.....

have a great cyber Monday....the cold wind is howling this morning.... hope the sun is out...🤔🤔😊

PowerJack comes in a different version (either upgrades or downgrades). have not tested all of the parts machine just yet >>> but it will be a back-up inverter...

I have not power on my 9000 kw with the inverter switch  yet  but all the mosfets are good .   The ribbon cable has one wire disconnected .  The AC output board has a broken Thor  connector .    Look like the worker  miss wire the AC output board .      I  have to bench test the LF driver  as I  do not want to  blow up the FETs  and destroy the  main board  worth 200  dollars .     That  why  Sean do not  sell  parts from a working  inverter  as the  parts may not work on a  similar  Powerjack .      Unfortunatly  the  rev 11.3 control board  will not replace the rev 11.0 in your 20kw  but  will replace the rev 11.1  in my 9000kw if tested bad .     

That's known issue with EG8010 even for LF inverters using EG8010.  I believe the Backshed inverters using EG8010 often disconnect the overcurrent feedback to avoid nuisance trips.

That  issue  will blow up all the FETs  when the nuisance trip occur when running  a heavy inductive load .      The  massive voltage  spike on shutdown  kill all the FETs  according  to test  with the  OZinverter .    I  stop testing EG8010  when I am ahead and only  use for light bulb .   

Thanks for making this thread has been really interesting to read.