I ordered the cheap 12k from powerjack and got something else

[Steve]

Both my units have the choke warning

I blew up the upower starting the well but they sent me the parts so it runs again 

[RobertM]

3 hours ago, Steve said:

Both my units have the choke warning

I blew up the upower starting the well but they sent me the parts so it runs again 

I'm going to open up my PJ AMG 8000 to read the FET value. Your inverter has one more cap, and I'm curious to see if they put "better" FETs in your unit. People tell me caps can be added to these main boards, but maybe higher level MOSFET boards are available, too. The regular MOSFET boards for my unit are $25 ea.

[Sid Genetry Solar]

36 minutes ago, RobertM said:

I'm going to open up my PJ AMG 8000 to read the FET value. Your inverter has one more cap, and I'm curious to see if they put "better" FETs in your unit. People tell me caps can be added to these main boards, but maybe higher level MOSFET boards are available, too. The regular MOSFET boards for my unit are $25 ea.

I will note, that no chain is stronger than its weakest link.

• For continuous loads, the undersized transformer is the issue
• For surge loads, the poor MOSFET drive is the issue
• For loads that cause a red light and error, the control board / firmware is the issue

Putting more expensive FETs in a PJ inverter will just make the smoke more expensive. 

But seriously now, if there's 6x FETs/board, they're likely rated somewhere near 100A each (at least that's the TO-220 package limit!)  So 600A * 24v = 14,400W should be a safe continuous maximum for the FETs (not sure about the internal wiring or PCBs though!)  14,400W / 240v = 60A on the output (not counting losses).  Obviously, however, the inverters can't reach this without blowing up.  My point is that the FETs aren't necessarily the issue.

For example, it doesn't matter how many horsepower your gas engine is rated for, if the spark plug wires are shorting out and the timing's wrong!  It'll run very poorly, and maybe even blow up if pushed hard.  But that doesn't necessarily indicate a bad engine, as much as a poorly-controlled one.

[RobertM]

34 minutes ago, Sid Genetry Solar said:

I will note, that no chain is stronger than its weakest link.

• For continuous loads, the undersized transformer is the issue
• For surge loads, the poor MOSFET drive is the issue
• For loads that cause a red light and error, the control board / firmware is the issue

Putting more expensive FETs in a PJ inverter will just make the smoke more expensive. 

But seriously now, if there's 6x FETs/board, they're likely rated somewhere near 100A each (at least that's the TO-220 package limit!)  So 600A * 24v = 14,400W should be a safe continuous maximum for the FETs (not sure about the internal wiring or PCBs though!)  14,400W / 240v = 60A on the output (not counting losses).  Obviously, however, the inverters can't reach this without blowing up.  My point is that the FETs aren't necessarily the issue.

For example, it doesn't matter how many horsepower your gas engine is rated for, if the spark plug wires are shorting out and the timing's wrong!  It'll run very poorly, and maybe even blow up if pushed hard.  But that doesn't necessarily indicate a bad engine, as much as a poorly-controlled one.

OK, then. I seem to remember Ray Jungle Power on Ebay claiming  (before he vanished) that PJ used only top quality western transistors. I'm making progress towards another pump test.  I'm connecting the super capacitor bank in a better way. I will use 8 gauge copper cable, which is about the diameter of the terminal holes and just solder as securely as possible. It consists of 12 350F in series. I will charge it to the same voltage as my combined lead-acid and LiFePO4 (~ 20A at 230V) battery bank. By my calculations, the supercaps should provide an additional 2.5A @ 230V boost for 5 seconds.  My guess is that, in the earlier test, the inverter shut down due to voltage sag.

It pumped for a few seconds, then shut off with a red light.

 

[Steve]

A few points on this

My upower is a 24v unit with 80A fets 6ea

That's 11.5KW if it was driving a square wave as a surge load if everything was perfect

That makes the fets a definite weak point for surge loads.

The ASL2 transformer doesn't dissipate heat well limiting the continues load but not the peak.

Poor drive will cause poor efficiency as the junction resistance will be higher than it should during transition time 

 

Edited July 31, 2022 by Steve

[Sid Genetry Solar]

1 minute ago, RobertM said:

OK, then. I seem to remember Ray Jungle Power on Ebay claiming  (before he vanished) that PJ used only top quality western transistors.

Hahahahahaha, teeheeeheeheee, marketing 101 🤣😁🤪

At least the past few years, they've only ever used Chinese FETs.  Huayi (i.e. HY3810), NCEPower (i.e. NCEP039N10) and Ruichips (i.e. RUH1H150R).  Never seen an "imported" FET in the inverters.

Now, that's not necessarily a bad thing.   Some Chinese FETs are definitely garbage.  But we haven't had any FET-related problems with NCEPower or Ruichips (as being the only ones we've tried).  My house inverter currently has Huayi HY3810 FETs in it.

FWIW I did try some Huashuo FETs, and those things were a solid definition of garbage.  Flimsy weak, soft leads...and the manufacturer was ashamed to even print their own logo (or a datecode) on the package.

[Steve]

Another point

Motors generally run at a reduced power factor

So 10A resistive load PF1 is 2400W @ 240

but 10A motor load is likely only 1680W due to the PF of 0.7

the FETs will still deal with the 10A load ÷ the duty cycle so a 10A load at .7 has a FET current roughly equal to a 12A resistive load 

Edited July 31, 2022 by Steve

[RobertM]

2 hours ago, Steve said:

Another point

Motors generally run at a reduced power factor

So 10A resistive load PF1 is 2400W @ 240

but 10A motor load is likely only 1680W due to the PF of 0.7

the FETs will still deal with the 10A load ÷ the duty cycle so a 10A load at .7 has a FET current roughly equal to a 12A resistive load 

 My well pump is a  230V 3 wire pump (2-115V hot, 1 neutral, 1 ground) rated as 1/2 hp. The locked rotor amps (inrush) for 1/2 second is 20.5A @ 230V.

Edited August 1, 2022 by RobertM

[Steve]

You need to be capable of 25A for a few seconds at the FET level 

[RobertM]

1 hour ago, Steve said:

You need to be capable of 25A for a few seconds at the FET level 

The lead-acid battery bank alone has 1050 cold cranking amps at 26V. That would be over 100A at 230V. The LiFePO4 bank supposedly has a 100A BMS.

[Steve]

The battery will never see the FET pulses because of the caps

The way the sign wave is made is through a wave slicer (class D amp)

The instant voltage is generated by what % of the time tha FETs are on

If the peak current is at the max voltage the FETs are on for most of the time but inductive loads the current lags the voltage so as the voltage is decreasing the current continues to rise causing the FET on state to hold larger currents for shorter duration 

[Steve]

Another note my pump only pulls 9A running but my clamp meter shows over 60A inrush

The inrush is much higher than the locked rotor current 

[Steve]

It's possible you could add a cap across the motor increasing it's PF making it more inverter friendly 

[RobertM]

1 hour ago, Steve said:

Another note my pump only pulls 9A running but my clamp meter shows over 60A inrush

The inrush is much higher than the locked rotor current 

I thought inrush meant the same as LRA.

Edited August 1, 2022 by RobertM

[RobertM]

1 hour ago, Steve said:

It's possible you could add a cap across the motor increasing it's PF making it more inverter friendly 

The motor is submerged, 50 ft below ground. I may put a 50uF choke on L1 and L2.

 

[Steve]

Caps and inductors shift current phase

The motor is an inductor so current lags voltage

capacitor current leads voltage

Putting a capacitor across a motor shifts the current phase back and if correctly matched can get it near 1.0PF actually reducing the run amps

I'm not sure how it would effect inrush 

[Sid Genetry Solar]

7 hours ago, Steve said:

Another note my pump only pulls 9A running but my clamp meter shows over 60A inrush

The inrush is much higher than the locked rotor current 

I don't see how the start current could be higher than the LRA nameplate rating.  For a 9A loaded running current, I would expect a much higher LRA than 20A...the 60A you measure sounds about right.  LRA = locked rotor amps = the absolute maximum power the motor can draw.  (The only way to run the current higher than that would be to purposefully drive the motor backwards.)

An itty bitty 1A running current refrigerator compressor will easily have an 8-9A LRA rating.

[RobertM]

1 hour ago, Sid Genetry Solar said:

I don't see how the start current could be higher than the LRA nameplate rating.  For a 9A loaded running current, I would expect a much higher LRA than 20A...the 60A you measure sounds about right.  

 

I found specs from the motor manufacturer. My motor is a 1/2 hp 230V "premium" motor, the second one in the chart.

[Steve]

Looks like your pump is about half the size of mine

I wouldn't expect an issue starting it

[Sid Genetry Solar]

There has to be a fatal miscalculation in that specsheet.

The DC resistance of the "Main" winding is listed as between 5.1 and 6.1 ohms (presumably ohms??)

230vAC through a 5.1 ohm resistance = 45A.

I don't know if the start winding is always cap coupled, or switched--but regardless, the start winding is listed at 12.4 ohms resistance.  At 230vAC, that's 18.54A.

So if we add both winding amperages together, you get (45 + 18.5) = 63.5A LRA possible.

 

...which lines up with the 60A you measure.

[RobertM]

1 hour ago, Steve said:

Looks like your pump is about half the size of mine

I wouldn't expect an issue starting it

I have trouble starting a refrigerator, and a 1/2hp 115V motor unless a small load is already running on the inverter. I didn't know this trick until recently, and haven't tried it with the well pump.

[Steve]

That is a logic board issue

My Upower dropps out for a sec starting just about any motor loads but the Plus Energy doesn't

I was kinda thinking it could be related to lower DC surge current due to higher input voltage 

[Sid Genetry Solar]

1 hour ago, Steve said:

I was kinda thinking it could be related to lower DC surge current due to higher input voltage

Surge ability will actually be HIGHER with higher input voltage...as even with the same amperage limit, an increase of voltage (volts * amps) will result in higher wattage.

[Steve]

That was my logic

Just wasn't sure if the power Jack logic boards compensated for it 

[Sid Genetry Solar]

7 hours ago, Steve said:

Just wasn't sure if the power Jack logic boards compensated for it 

Ha.  No, nothing of the sort.