Author Topic: guess who bought a power jack inverter  (Read 204785 times)

0 Members and 2 Guests are viewing this topic.

Offline oztules

  • Moderator
  • Hero Member
  • *****
  • Posts: 1177
  • Karma: +105/-8
  • Village idiot
Re: guess who bought a power jack inverter
« Reply #180 on: April 03, 2015, 02:39:38 pm »
You have replaced the transformer... so we have to assume it is not that.

The transformers can also act as a loud speaker, telling you what signals are going on inside through it... it is after all a coil without the cardboard cone, but it still makes noise.

It is possible that the control card has either a loose driver ic, noisy totem transistor, or some other temp sensitive problem, introducing spurious noise into the output stage.

Replace the control card with the other one, and see if that stops it... if so warranty  if it is new, or buy another card.... we need to  isolate the problem .. this may do it.

It could be a broken/cracked track on the power card going from the ground plane to the fet source... some of those tracks are pretty thin... use a magnifying glass to check this.... and a bright light.

It is hard to see other problems as being on the power card, unless you have a poor connection anywhere there... check bolt tensions  on lv leads both on the sinks and the inlet terminals.

I would have expected loose terminals anywhere to show up by now as heat spots, so noisy transistors on the control card is the  most probable source of traceless noise that is not the transformer......also check your inductor.

In closing, it is a modular unit, so you can replace the control card, or the power card or the transformers one at a time to see where it is coming from. You may get some result by measuring the output voltage of the power supply on the board.. is has a 3846 type chip in normal pwm mode supplying the voltage for the units operation and high side supplies as well... any noise in this circuit effects everything.

The beauty of these things is that you can source all the parts cheaply, and have spares on hand for peanuts.... unlike the name brands where a fault is a real headache/heart stopper ( ask Rossw ), and you won't have spare boards.... and it will be an expensive expensive problem to deal with.... at the very worst, this thing will cost less than $200 to completely rebuild.. less the transformers.... thats amazing!!

Voltage increase:

The voltage divider on the cards, is a very high impedance ladder of 4+megohm resistors ( from memory).... and is very susceptible to moisture if you have no anti wetting agent there in a damp environment ( read CRC or equivalent). Any ( even megohm plus) leakage between rails in this ladder will effect the divider voltage the chip sees.

So even a 10meg resistor placed across the ladder at different points from the ac source down to the eventual chip divider will make an increasing difference to the output voltage... it is a long divider network.... very substantial difference would be found by shunting the  last 3 or 4 resistors in the strings with the at least10meg I would think.

Have not needed to do this, so I can't say exactly how to place the resistor for your application, but that is how I would approach it... or 1 meg for starters in series with the string should alter it a bit too.... but that will mean messing with the surface mount stuff invasively.

................oztules
Flinders Island...... Australia

Offline Wolvenar

  • Senior Moderator
  • Hero Member
  • *******
  • Posts: 1474
  • Karma: +40/-0
  • Mr. Murphys pawn
Re: guess who bought a power jack inverter
« Reply #181 on: April 03, 2015, 05:37:41 pm »
OK, is there a transformer, and a toroid in this?
Or did you replace a transformer with a large toroid wound to purpose?
Trying to make power from alternative energy any which way I can.
Just to abuse what I make. (and run this site)

Offline lighthunter

  • Sr. Member
  • ****
  • Posts: 323
  • Karma: +13/-0
Re: guess who bought a power jack inverter
« Reply #182 on: April 04, 2015, 01:34:51 am »
1st of all Thanks a million for your ideas and suggestions! It sure helps think about things I might have overlooked.

When i came home from work today, much to my surprise the system had run nonstop without any faults.  (for right now, setup depends on this wounded inverter for dump load which had to run most of the day!) 

Wolvenar, Single transformer unit came fused at20A i think, its approx a 3KW toroid. I replaced it with a used toroid i picked up on ebay for $100 this winter made in california by powertronics. It had specs of prim:37V/60hz sec 235v@12.5A. I put Em side by side and connected both to 120v and measured the original at 22.01VAC and powertronix at 21.83VAC. The powertronix is a fair bit heavier and larger diameter but not as thick. Anyway, it was close enuf for me to try it, (I can c a 240V/120Vsplit phase mod comiing in near future:). Functionally it was perfect replacement but I couldnt fit lid on of course.
Health Warning: May contain traces of nut!
LH

Offline lighthunter

  • Sr. Member
  • ****
  • Posts: 323
  • Karma: +13/-0
Re: guess who bought a power jack inverter
« Reply #183 on: April 04, 2015, 02:26:46 am »
cont.  So back to my surprise that wounded pj had run well all day, i listen close and problem still exists but not bad, i grab dc current clamp and check, 24v current is only 2 amps high (31.4A@700w) I hadnt had time to do any board swapping yet as Oztules suggested (thankyou). It was obvious that leaving the chassis open had improved the situation (come to think of it ive never seen fans come on on this unit). Anyway, from info Oztules suggested, and past bad  experiencs with dip sockets, i pressed down on driver IC's while running, the unwanted noise went away and current went down 1 amp. My other identical pj runs same load at 29A, the other amp comes from core size difference in toroid, i noticed idle current was up a bit immediately after changing transformer. Both inverters, idle at about 26w with original parts according to my test equip and procedures.(errors are inevitable, ex. dc v variance etc.

The noise, (originally was loud arcing sound with occasional pops) now is barely audible with occasional tick sounds will completely go away when applying pressure to A3120 driver chips and sockets but slight evidence of problem returns randomly.
Health Warning: May contain traces of nut!
LH

Offline lighthunter

  • Sr. Member
  • ****
  • Posts: 323
  • Karma: +13/-0
Re: guess who bought a power jack inverter
« Reply #184 on: April 04, 2015, 03:51:24 am »
Cont...  so it looks like i need to order a new control board. Now to decide if i want to go with 48v :)

In case anyone is wondering why im using 3 inverters and a 300lb transformer for a 2.2kw panel, its because of a very frustrating grid tie inverter im using. HF pure sine. Does a wonderful job at very high efficiency as long as you can satisfy it. Specs are, 3500w 2 channel mppt 13.5A each, dc range 100-600VDc mppt range 150-420. 208/240 60hz. for a while it seemed like this thing would go offline if you sneezed. This is a prodjguct of dow chemical (we want to look green) business venture with Emerson made for "solar shingles" which entered the residential market for 3-4 yrs and disappeared. The emerson grid tie doesnt like it when you switch loads connected between it and pj. The control starts out day by running a 360w 230v to 24v led ps to charge bank until sun is strong enough then Pj turns on, 5min countdown starts, @ T=20 sec a medium load starts, once grid tie comes online and finishes ramp up then load will either stay on or turn off depending on battery level and solar input. As long as power removed  gti seems ok with it. The GTI also doesnt mind if pj is putting away power into batteries at levels of 2150w. It just doesnt like to be tickled by loads switching. I ordered some zero crossing SSRs that may solve issue. It doesnt help to have 300lbs of transformer core between GTI and pj, im guessing the abrupt change in current when a load is switched produces a spike, no voltage drop under load though (10GA square wire 240v center tapped winding). Therefore any xtra power over approx 1kw must be sent to bat bank and taken out by second inverter. Great sun lately though, 42kw in 5days with many 5min resets during first 4 days. Hopefully solid state relays will fix problem, ive tried moving 240v grid tie point up and down 10 v seems to run best at 245 tap but this can crawl to 259 when producing heavy. Limit is 266. this emerson gti was not made under AS4777 only i think it had to comply with another code for us rules, iwhile the pj is backfeeding power into battery bank the needle on amp guage seems to be indicating a pulse every second that is positive going doesnt seem like skipping cycles, it checks for a grid in about every imaginable way. It will be a couple of days before i get back to this but thanks again for all the help!
Cheers, Lighthunter
Health Warning: May contain traces of nut!
LH

Offline oztules

  • Moderator
  • Hero Member
  • *****
  • Posts: 1177
  • Karma: +105/-8
  • Village idiot
Re: guess who bought a power jack inverter
« Reply #185 on: April 04, 2015, 03:15:26 pm »
Well looks like you have narrowed it down a bit.

May be a noisy driver chip too, perhaps swap them around and see if the problem follows one of them.... may also be a fracture in the fine tracks in the board.. pressing down the chip affects more than just the legs to socket interfaces.

Zero crossing will be worse if your driving/switching larger transformer loads... they hate zero crossing start... opposite to what you would think.. they prefer full peak start, they will then exhibit no inrush current... at zero start, they will heave with inrush currents  for the first few cycles until the magnetization catches up with the program... inrush on torroids can be debilitating, as the current can be 60 times magnetization current. Just built a power supply with 3kw torroid... needs a soft start circuit, or blows the 20amp overloader instantly about 50% of the time.... but once running has very low current at idle.

Resistive loads will do better though.... nothing is straightforward.... sigh...

The quality of the silicon steel used in the torroids will strongly affect the idle current. If the PJ tranny is heavier on the current, then it is lower quality than the other, or if the californian one is drawing more, then it is lower quality... depends where they get their steel from, and at what grade.... assuming same wattage ratings. Idle temp will give it away too.

In this country, high grade transformer steel is hyper expensive, so there is a cost analysis done very carefully... and a compromise will be reached.... I suspect in China, the difference will be less, so they may opt for the better steel as the price may not be that different..... it seems volume makes a difference to everything.

I got some transformers from the HEC ( Hydro Electric Commision), and the EI transformers were 4 times better idle current than the run of the mill laminates with eddy currents etc... like buying them from Nasa I guess.... money was no object.

Back in the day, I got hold of some racking units from the Goddard space flight center at the hyball project in Mildura for the 60's moonshots ( upper atmosphere weather up in the dessert) It was a Nasa thing.... all ......and I mean all ..the wiring was pure silver, and teflon coated...never even heard of teflon then... (ever tried to strip that wire???)....no copper anywhere to be found..... nice to have deep pockets. They were built to last forever...but  probably outdated by the time they were pushed into operation though.

Always 48v for me.... never less.

............oztules
Flinders Island...... Australia

Offline lighthunter

  • Sr. Member
  • ****
  • Posts: 323
  • Karma: +13/-0
Re: guess who bought a power jack inverter
« Reply #186 on: April 05, 2015, 04:40:11 pm »
 :) You are absolutely right Oz, switched the driver chips between 2 inverters and now the noise is in the other inverter, accidently put cover on b4 checking it out and the heat thrown from toroid really causes trouble. Lid off, fan on chips, works ok. (New chips are ordered, A3120 opto).

I also did the split phase 120/240 mod. Works great! Except, MR. GTI wasnt impressed. Wouldnt even acknowledge AC was connected should have been within parameter limits oh well, so I get out the utility knife and cut a trace, jumpered in 470k resistor and PJ went from 111v to 116. (GTI parameter say 107min 130max phase to neutral) Turn power back on and believe it or not, i can add and drop 110v loads without a hitch I even hooked up 800w water heater element hot with arcs an sparks, the grid tie never even flinched. It wont let me connect loads to 240v side though. Now i need more batteries and (driver ICs of course). Project is close to done. Thanks Oztules and Wolvenar for thoughts an ideas! To increase PJ output voltage cut trace between R259 and R262, solder in A resistor to join back up and you are good to go. 470k gave me 5 volt increase.
Later Lighthunter
Health Warning: May contain traces of nut!
LH

Offline off the wall

  • Full Member
  • ***
  • Posts: 76
  • Karma: +3/-0
Re: guess who bought a power jack inverter
« Reply #187 on: April 07, 2015, 03:45:54 pm »
I had a bit of a bright idea . . . which turned out puzzling.

Ages ago I bought a cheap ebay LF unit from Power Jack seller Albee and it turned out to be one of those nightshift units - sold as a 6000W unit in a big box it only had one transformer in it with the rating blacked out. It served a time but it was the first generation of power boards which I found unreliable, and it blew its FETs one day for no accountable reason. The boards have been upgraded so I guess it was a common problem. I got the new power and control boards to fit but didn't get around to doing so.

I'm based on a 24volt system so everything that I write currently is in respect to that. (I've got a 48V system planned . . . )

Interestingly the transformer is labelled 17.5V primary and 230V secondary.

Having been given some defunct UPS inverters I've tried the transformers from those on a spare set of PJ boards . . . and when connected on the low voltage side, the transformers not connected to the control circuits on the high voltage side gave between 220 and 250V depending on the transformer, but the electronics shut down within around 10secs sounding the alarm, presumably because it didn't detect the secondary voltage. When connecting the secondaries to the boards as intended, the output voltage was only 180V to 210V depending on the transformers, so using these transformers with the PJ electronics was an abortive idea.

I'm using a PJ LF 8000W to power part of my house off grid. It's great but we have to remember not to turn the kettle on when the washing machine or dishwasher is on, a significant disadvantage.

This unit uses two transformers. The primaries are in series, as if capable of serving the 12V range of units, with the secondaries in parallel. I had the great idea to use the transformer of the 6000 - 5000 unit, also 24 volt in parallel with these two.

However, when running the 5000 transformer in parallel with the two transformers of the 8000LF it gives only 215V whilst the two in parallel give 230V. Putting the output in parallel with the 8000LF transformers causes internal consumption of 1250W with no output.

This seems rather high for a mismatch of only 15V at the output but clearly to be successful I need an exact duplicate 8000LF to double up.

It would be nice if there was a way to get control boards to synchronise so that one could piggyback these inverters.

Best wishes

OTW

Offline oztules

  • Moderator
  • Hero Member
  • *****
  • Posts: 1177
  • Karma: +105/-8
  • Village idiot
Re: guess who bought a power jack inverter
« Reply #188 on: April 07, 2015, 05:06:47 pm »
"I'm using a PJ LF 8000W to power part of my house off grid. It's great but we have to remember not to turn the kettle on when the washing machine or dishwasher is on, a significant disadvantage."

If you know you will not drive it too hard for the transformers, the current bypass around the CT will solve the coffee dilemma....  but if you drive it too hard for too long, then your transformers will over heat.

Mine had bypass for 8kw cont, bu the transformer is a monster too.

You can alternatively lower the resistance of the resistor next to ZD2 at the front of the board where the 12v goes off to the display It has the words adjust next to it, and looks like a hand solder, not machine solder in general. It is a shunt across the output of the CT, and so will lower it's EMF and so allow the current to be higher than the 12 secs or so.

BUT this should not be done to run the thing for extended periods of 8kw... your transformers would need oil cooling for that, or bigger ones..

Also if you intend to drive it harder for longer, you will need a better cooling system, or run the fan continuously works too. This is due to the current seen by the computer is not what is really happening. The fets don't mind in the 48v units, but the fan algorithym is messed up more than it is norma;;y... I don't use their fan at all now, have my own controller in there.

Your 15v mismatch is enormous for torroids. The reason these things can run so well for so small a size ( compared to EI units) is their very tight magnetic properties... ie there is very little leakage, so 1volt mismatch would be absorbed in the leakage fairly readily in an EI unit, but in these things it will make a real difference.... 15v will make a huge monumental difference.... and I am surprised the idle power did not get worse than 1kw. You will need to add turns and get them the same before you attempt to parallel them.... you may be able to fit a few turns of battery cable in there to equalise them.

If you wanted to synchronise them I would intercept the control signals to the totem and the optos,and get a take off, then  isolate them on the second board, and feed them from the first.... common the ground, and it should work. It means on the second board, none of the control functions will have any real effect, so the first board in fact controls everything, and only uses the fets and drivers of the second board... the second board will see the output, and try making correction perhaps but will have no effect, but the fans will come on in high temp only, not current...... very crude but would work I think....... also the current sense will only reflect one output... but this would be a good thing in this case.

Pleased you have nailed it Lighthunter.


................oztules
Flinders Island...... Australia

Offline off the wall

  • Full Member
  • ***
  • Posts: 76
  • Karma: +3/-0
Re: guess who bought a power jack inverter
« Reply #189 on: April 07, 2015, 05:23:44 pm »
Thanks

Very helpful indeed.

Yes - I suppose I could add turns on the secondary of the 5000LF transformer that's low by 15V

I wonder if anyone's looked at those transformers to find turns per volt on them? Didn't I read somewhere that someone had rewound a toroid or was that from one of the scrap grid tie inverters?

Silly question - but there's no point in reinventing the wheel - has anyone drawn out the circuit diagram surrounding the optos to isolate and run from another board? But in fact the result would still be the same with mismatched transformers as they require the control signals dedicated to each transformer to regulate for the 230V output.

Many thanks

Best wishes

OTW

Offline robs

  • Newbie
  • *
  • Posts: 1
  • Karma: +0/-0
Re: guess who bought a power jack inverter
« Reply #190 on: April 07, 2015, 05:35:27 pm »
Hi everyone

Bought this inverter from Power Jack 5000 LF couple of months ago but never got to actually use it until i got the battery bank ready. got that ready, got solar panels to charge my battery but my inverter stops working for some reason. I think is because it overheats, trying to use my washing machine and when the water heater from the washing machine kicks in it gives out 3 short beeps and turns off. The watts on the display goes up to 2500 - 2600 and i thought it supposedly halve a continuous 5000 watts.  Any help is appreaciated, hence i can't get in contact with the seller from ebay.  Thank you

Yes it's all true.
In a moment of weakness I decided to buy a power jack inverter to play with.... and it looks like this.
(Attachment Link)

It looks like this inside... warranty blown right there......

(Attachment Link)

(Attachment Link)

(Attachment Link)



Well there it is.
This one is a 24v unit ( my mistake ), but it can do what it says. I can run loads of 6kw or more without problems.... I had the electric stove on, the hot water on and then the water jug..... and it turned off after a minute or so.... can't complain really, the stove was 3200w the hot water some 2500w, and the jug another 2500w... so all up it tried to do over 8000w... I don't now if it was batt under voltage, or power overload... one or the other, because over 8000w is over 300 amps, and I think I had voltage drop in the line... surprise suprise.
The 200 amp meters were way past their stops, and it will never again have to try this, but i could not get it to trip any other way.

So it was impressive to say the least. It also started the 3 phase converter, which is a 3 ph 10hp induction motor running in delta@240v from single phase...... so it can start darn big inductive loads too..

About this time I was becoming rather impressed, as i didn't think it had the  ticker for this kind of abuse, so it seemed I needed to take a more serious look at this thing.

I had bought it because it was darn cheap, with free shipping to this island.... and as it weighs over 60lbs or 32kg, it is not so easy to get stuff that heavy sent here... so I bought it as a back up.

I was becoming of the view that this wasn't a toy like the ones I had heard and seen on the web earlier.... so decided to do some more testing.


Well it wasn't all tea and bikkies I'm afraid.

The engineers have not addressed the idle current other than a sleep mode at 25 watts.... but an awake mode of no power draw costs a whopping 8 amps..... yes you read that right .... 8 amps at idle.

Now as a cheap back up this is acceptable.... but as a house driving unit.. it is not good news to burn up over 5kwh a day and do nothing.

So it was time to try a few things we have learnt from the other unit I have ( the power star thing).

It already has two whopping torroids ( worth buying just for them ), so it was not going to be addressed like the power star was.

So I wound this:

(Attachment Link)

This solved the problem.... now idle currents were down in the sub 1 amp range.... in fact now it could be alive all the time, and waste less than 20 watts..... less than their sleep mode...... so 3 and a half turns around a 65mm e core made the thing very very acceptable as a front line unit..... who'd have thought.... a power jack as a front line unit.

Well it's true, it has been running the house... including the electric hot water for the last few weeks. The only thing on grid is the stove.... so usage from the grid was about 1kwh per day, and the house was running at about 11-13 kwh/day.

There was one other thing of interest here.

I have panels running at 300 volts and more intp grid tie inverters into the batteries via the inverter.... ie the grid ties run on the output of the inverter. When the house loads are greater than the grid ties can manage, the inverter takes up the slack, when the grid ties put out more than the house can use, the extra is fed back through the inverters output, and back through the sitching fets and transformers into the batteries.

I have seen in excess of 170 amps being back fed into the batteries via this method. The drawback is you need a big dump load on the batteries, or a cut out system to sequentially turn off the grid ties to keep the batteries from over charge.

If this is allowed to go, then the power jack will turn off when the batteries exceed 32v... which is the final safety in all of this.....THIS DOES NOT INVOLVE THE GRID for those folks wondering..... this is strictly off grid antics.

It is interesting to note, that without the e core filter in series with the primaries of the transformers, it would not grid tie to the grid inverters... but rather, blow all the overloaders in the ac line to the inverters.

With the filter in place, it was perfectly happy to be the grid, and sync the grid ties to it's output, and reverse flow the power to the batteries.... but not without the carrier filter.

So the big filter did two things, made the inverter a classy unit, and allowed it to back feed grid tie inverter inputs into it's outputs.

To say that I am impressed is an understatement. It appears bullet proof, and the spare parts ARE available.

Thats the clincher for me, in fact after I chased them up, they have even put the spares on ebay... about 170 dollars for the entire innards of the machine... ie power board, and control board, battery selector board, and driver board for the display..... just add transformers and a box, and you have built your own..

Before you rush out and do just that...... it is heaps cheaper to just buy theirs, as the cost of those torroids over here are over 1500 dollars alone, cheaper in the UK, don't know about USA. The transformers are over 25 lbs each....

One problem I had was the 24v part... I was set up for 48v..... so I had to do something about it.
I rearranged the battery pack to reflect the new setup... now it was 1500AH @ 24v

Re wired the solar panels for 24v ( left 4kw as 350v for the real grid ties ), and that gave me about 130amps of 35v solar power.

Now I needed a controller.
I sat down and scribbled a circuit of sorts out, and fleshed it out as I designed the board.. and finished up with this

(Attachment Link)

It is about 10am, and already it is up at 80 amps

(Attachment Link)
I have seen it at 120 amps. The heat sink ( and no those burn marks are form when this heat sink belonged to someone elses 2500w HF pure sine inverter) stays at less than just over skin temp even at high currents. The wave form is very clean, but swithcing at only about 4khz... makes the "layout" simple.

Here is the current going to the batteries after the power jack has used a few up from the incoming 80 odd amps
(Attachment Link)

May do an article on that if folks really want to try it.

Here is the wave form of the makeshift "solar controller" ( not the power jack.... that  is pure sine) when it is running at only 40 or so amps after the bulk of the charging is done ( mid morning )
(Attachment Link)

All in all...... I love my power jack..... did not expect to ever say that in a million years.


................oztules

Offline oztules

  • Moderator
  • Hero Member
  • *****
  • Posts: 1177
  • Karma: +105/-8
  • Village idiot
Re: guess who bought a power jack inverter
« Reply #191 on: April 07, 2015, 05:36:45 pm »
I rewind the torroids from the grid ties, and until you stack them they are in the 1.14v/turn... would guess those PJ are close to 1v/turn... but have no real idea.

If you really want to play around a bit, you can improve the efficiency of your present transformers.

You need to unwind the secondaries, and stack the two cores with their primaries together as a stack, and rewind the secondaries, with half the turns and twice the cross section..... you will end up with lot pf spare wire, as you will have gotten rid of 2 x coil ends in the process, so less resistance.

If you really want to drive the thing hard and use the jug all the time with the washer, and the dishwasher, you need to totally unwind the three transfromer cores you have, and stack them all together, epoxy them perhaps to keep noise to zero.

Then rewind the primary on the three stack at 1/3 the turns of the orginals, and then the secondaries at  1/3 the original turns and three times the cross section, then you will have eliminated effectively three x two coil ends on both the HV and the LV side.

Your resistance will have come down significantly, and your power handling has gone up.... plus the more you stack, the easier it is to wind as the turns drop off quickly...

So thats something to play with if you can't get grid ties easily.


Yes, the transformers need to be matched for stacking... use two identical PJ units..... easier to modify and use more transformers.

I now have the 15kw driver boards, so they need no mods for the 8kw cont or thereabouts... but no different than fooling the 8kw ones in reality, as I don't use their fan regime.... the transformer is bigger than most old industrial welders though


................oztules
Flinders Island...... Australia

Offline oztules

  • Moderator
  • Hero Member
  • *****
  • Posts: 1177
  • Karma: +105/-8
  • Village idiot
Re: guess who bought a power jack inverter
« Reply #192 on: April 07, 2015, 06:11:26 pm »
Rob, the 5000w from PJ will never get there unless you modify it. They are generally good for about half the rated continuous power... what we call continuous, and they call continuous differs remarkably.. for them 12sec is continuous, for the rest of us it is indeterminate.

There are a lot of variants of the PJ it seems, you never get two the same, and the pricing seems to change what you get. If you win a cheap auction, the machine may have smaller transformers, and less transformers.... and less continuous power... they seem to build whatever, and program it respectively.

It can be made to do it, but requires modifications to the transformer, and the current transformer on the board. The current bypass around the CT will get you your power for sure in a few minutes of work..........., but your heating will be problematic.... they are cheap for a reason, but are brilliant value none the less.

That quote is from my first PJ... I bought it fairly dear, but it had two of their biggest transformers in it, and it was programmed accordingly.... very impressed I bought another cheaply on auction not buy it now  expensive.... it had small transformers and worked accordingly... thats when I realized there were PJ's and PJ's, and we get what we pay for to a fair extent... it is still a lucky dip i think.. but dollar for watt they are superb, and you can get the parts if you need.

If you want more power, and have no electrickery bones in you, then there is only the option to buy a much bigger unit, and allow for about half power continuous if PJ
The other alternative is to get the W7 powerstar units, around the same sort of price but they will do what they say, but you lose 5kwh/day in idle costs you can't change without changing the transformers to torroids or high quality laminated EI transformers.

There are very few cheap LF inverters out there. The flagship ones like the selectronic or Latronics are very expensive.. so not on my radar, the victron LF hybrids are expensive and have no guts for driving decent sized 3hp motors.. so we are left with PJ and whatever mods we think we need for our purposes. At least the PJ can be anything we want, and parts are available for those of us in remote locations.

If your in Aust and can get inspire grid ties for peanuts, then a rewind and a CT bypass would get the 5kw up to continuous for a hundred dollars or so. If you have no such inclination, resell the unit and buy something that you feel will do what you want... but it is not hard to  make these things do anything really.

What voltage are you running?

...............oztules
Flinders Island...... Australia

Offline off the wall

  • Full Member
  • ***
  • Posts: 76
  • Karma: +3/-0
Re: guess who bought a power jack inverter
« Reply #193 on: April 08, 2015, 06:18:57 am »
OZT - thanks for the info about stacking the transformers. Very very interesting. I'm sure that I might not be the only reader to be a little confused by your description as to what remains wrapped around what and which gets rewound around the lot.

But my priority at the moment is a quick and dirty fix . . . ;-)

Certainly these units do very well and even unmodified this morning I found clotheswasher on, toaster, hoover, 750W dehumidifier and in all just exceeding 4kW for quite some time. One would pay a fortune for a Victron to do even that . . .

There are a lot of variants of the PJ it seems, you never get two the same, and the pricing seems to change what you get. If you win a cheap auction, the machine may have smaller transformers, and less transformers....
 . . . .  I bought it fairly dear, but it had two of their biggest transformers in it, and it was programmed accordingly.... very impressed I bought another cheaply on auction not buy it now  expensive.... it had small transformers and worked accordingly... thats when I realized there were PJ's and PJ's, and we get what we pay for to a fair extent... it is still a lucky dip i think.. but dollar for watt they are superb, and you can get the parts if you need.


I've realised that getting even two of these units exactly the same might be a challenge. I was offered a 6kW 12V unit at good value enough for me to keep in stock and lots of people want 12V units for mobile uses. Thinking that it would have 2 3000W transformers was a mistake - it has one transformer, two fans now, and the single transformer labelled 8000W. At 12V really this doesn't matter as people's cables and batteries will be the limiting factors rather than the inverter really preventing it working for long periods at high capacity. Anyway - I digress - I didn't get my 2 transformers there to put together for extra 24V capacity.

So for the quick and dirty fix. . . . I've taken the 5000W LF transformer out of the case and the coils are neatly wrapped in some clearly high temperature film. There are three options:
1. Possibly remove a turn from the primary
2. Add an extra coil of around 10-15 turns to put in series with the secondary to increase the voltage by 15V.
3. Do something more radical

2. will be easy but I could do with a source of high temperature film to keep the new turns in place.

1. could be easy but messy.

Does anyone know how the PJ transformers are wound? Primary on top of secondary or underneath? If on top removing a turn from the coil could be easy. I don't want to explore and make a mess of a very neat and well constructed transformer . . .

Best wishes

OTW

Offline oztules

  • Moderator
  • Hero Member
  • *****
  • Posts: 1177
  • Karma: +105/-8
  • Village idiot
Re: guess who bought a power jack inverter
« Reply #194 on: April 08, 2015, 11:48:45 am »
Torroids are remarkable things to work with .. everything about them is wonderful... except winding the things.

If you want to take turns off ....... without invasive surgery ( 240 winding looks to be internal winding) you need only wind a few turns in the opposite direction..... on the outside...... and you have it........ so easy with torroids. Same with the LV windings, but physically more difficult as they are lots in hand wire.... so they are very easy to "tune".... we pay a price of more copper and associated R, but in this case, it will be in the HV side, so not as critical as the LV high current side. I haven't messed with the PJ transformers, but assume they only run 2mm wire on the HV side.... so easy to wind turns on the outside and change things.... up or down.

So every turn in the opposite direction nullifies that many original turns, any added turns in the original direction, adds to the original winding..... so you can see we can modify without messing the original one up. ( this assumes we series the new turns with the original winding).... so try it, wind a dozen turns around the 5kw units tranny, and series these turns with the 240v winding.... first one way, then the other and see the difference. ... they add or subtract....... now thats quick and dirty fix...

The stacking is very useful too. Same idea. If you strip all the multi core winding off the outside, it leaves only the 240v inner winding on the core. If we have two of these the same,we simply stack them one on top the other.... totally separate from each other.. just physically piggy backed. If they are not the same we can "tune "them as above.

Now we simply wind our LV winding around both of them, treating the new twin stack core as a single entity....... we then parallel the HV windings for twice the power handling in the HV side, and we have a new transformer, twice the power, lower resistance than if we had two separate ones in the LV winding because the winding path is now shorter... where we have the two stacks touching each other.... there is no LV wire ... just the original HV wires.... so we have saved that by 2 times.
We now have 1/2 the turns on the LV winding, as it now encompasses twice the core area than original, so half the turns, and so we can now double the wire cross section to make up for that...... we get a more powerful transformer as we have shortened the wire path, so resistance  goes down, as does temperature.

We can strip off the HV windings too, and glue the two cores together, and we free up more wire length, ( where the cores are glued...no wire, so we have saved those parts of the turns from each core... we have to wind only half the turns now, as we have doubled the core cross section area, so less turns to tussle with ( doubled the volts per turn)... now we are getting a big saving is space and resistance... and easier on the bloke winding it too.... can you see how much wire gets saved when you three stack... starts to get seriously good... 4 lots of wire sides are taken out of existence.... thats an entire transformer worth of wire saved for the same result, so the R heat of a whole transformer is now gone... and thats I^2R heat too.... thats how my big one is wound... three massive torroids.

Remember, with torroids it is only the wire in the core space that does anything useful, all the rest of it just gets it back into the core for the next turn.... ie three out of four sides play no part in the transforming, only the inner wires count, the rest is just resistance we suffer... all the magic happens in the hole in the middle.

Will do a mock up tomorrow as demo if I get time, and it will become clearer.
Flinders Island...... Australia