Author Topic: testing the egs002 inverter board  (Read 147627 times)

0 Members and 1 Guest are viewing this topic.

Offline Phaedrus

  • Jr. Member
  • **
  • Posts: 28
  • Karma: +0/-0
Re: testing the egs002 inverter board
« Reply #480 on: November 09, 2017, 08:40:28 am »
Hi Oztules,
Been reading the discussions here and on the BackShed forum,and would like your advice on some issues I could not understand :
From Back Shed :
1)
Quote
The egs002 or eg8010 are different, as they have a much tighter dynamic range for their spwm.. so for the 240v /48v version of the 002 or 8010 it will be more like 9:1 ( 26v primary)
Or a 13v primary for a 24v system... or 18:1
Could you explain ,why you would take a primary voltage of 26v,when the battery voltage is 48v ?And whats meant by tighter dynamic,range ?

2)Are you still using the temp control pin in lieu of the current feed back ? If IFB will trip the system at 0.5v,isn't it too close to the noise floor ,that it will give false trips?

Thomas

Offline Phaedrus

  • Jr. Member
  • **
  • Posts: 28
  • Karma: +0/-0
Re: testing the egs002 inverter board
« Reply #481 on: November 09, 2017, 10:04:33 am »
Another small query :
Would it matter if the CT was placed on the primary side or the secondary side? AFAIK it should not matter as long the transformer turns ratio is known/
Since the voltage is so small from the CT,do we have to compensate for the diode drop of the rectification bridge on the CT output ?
I have a fundamental working knowledge of electronics,so I hope my questions are not too over the top  :'(
TIA
Thomas

Offline rossw

  • Senior Moderator
  • Hero Member
  • *******
  • Posts: 879
  • Karma: +35/-0
  • Grumpy-old-Unix-Admin
Re: testing the egs002 inverter board
« Reply #482 on: November 09, 2017, 03:09:57 pm »
Since the voltage is so small from the CT,do we have to compensate for the diode drop of the rectification bridge on the CT output ?

Because of the way a CT works (*CURRENT*), the output voltage should rise as far as it needs to, to achieve the ratio of *CURRENT* in the secondary circuit (including the diode and burden/load resistor), thus eliminating the effect of the diode Vf.

Offline oztules

  • Moderator
  • Hero Member
  • *****
  • Posts: 1177
  • Karma: +105/-8
  • Village idiot
Re: testing the egs002 inverter board
« Reply #483 on: November 09, 2017, 07:54:00 pm »
"Could you explain ,why you would take a primary voltage of 26v,when the battery voltage is 48v ?And whats meant by tighter dynamic,range ?"

We need to generate a 240v sine wave, and to do that with dc we need to remember that the 240v is a rms value ( root mean squared )... so we need to multiply by 1.414 ( root 2) to get the real peak AC wave form voltage.. ie 1.414 x 240=  340 volts or more. We need to overcome sag in the secondary and primary, inefficiencies, and have room for high and low power outputs. The dynamics of the system, is how much variation is required to get the job done. A static system has no change, a dynamic one has change. We need to go from  0 watts out to over 14000 or more.... thats a big range... and it all occurs from the pulse WIDTH modulation of the spwm wave. The fundamental wave is static, not dynamic, but the spwm is very much on the move.... dynamic.... this width must change dramatically to create the sine wave, then the widening of those pulses to drive ever bigger loads.... and back to zero again.... thats the dynamic range required.

Ie the wimpy little pulse width required for idle voltage is simply scaled up ( width) as we need more power from the same waves.( Why it can go over 14000watts is beyond me.... it is truly amazing)..... but you can't get there unless you have enough over head in the transformer windings and ratio. If the software peaks out at a certain pulse width, then thats where you stop.......regardless of the feedback. If the width range is larger ( more dynamic) then you can push harder.

So a say 48v/70v idea would yield a primary voltage of 32v... but this gives no room for driving it much over unloaded voltage... so we need some head room... so 30 volts works okj, 28v works better...... 26 is getting a bit to low for me, but fine for most units. I tend to use 28v  or thereabouts. My transformers are tight enough, ( leakage and synchronous losses), that that extra few volts is enough to drive well over 60 amps@ 240v in the secondary, and into a 5hp induction motor for start up... it is probably a bit more than that, but my current meter is a little slow to catch it I suspect.


Put another way.........

The Power jack seems to have the software with a wider dynamic range ( zero to something higher than the 8010 in width)   , the 8010 has a lessor range. To make up for this, we use an even wider turns ratio, so that it takes less pulse width to get a higher voltage under load..... now we can see why we need a much lower "voltage" primary than 48v. A 48v winding just won't get anywhere near the voltage you expected.... ie 48vac  rms is actually near 70v peak...... putting 48v into a 70v system will not get the job done at all.... So 48vac rms  gives 70v peaks to work with the transformer, and 48v dc gives..... well just 48v...... so the output would be in the 160 volt rms range... with 240v peaks....... if we have a sine wave output....not what you expected is it? A normal AC voltmeter will read that as 160vac

Now..... if we had a square wave , then we would use a normal 48:240 transformer... thats 5:1 so our 48v gets to 240v rms just fine.... but it is just a square wave. So your transformer gut feeling was correct, but only for square waves from a switched DC source..... going from switched DC to sine waves is a bit different than the old multivibrator, now you need to look at the peak voltage required to get an equivalent AC RMS voltage.... suddenly were back to 340v  peak again.... minimum ( losses need to be looked after too)

The current feedback in the boards I left the gerbers for, use both the IFB and the 2110 shutdown pins.

Ross explained the CT business as well.


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

Offline ClockmanFrance

  • Full Member
  • ***
  • Posts: 171
  • Karma: +11/-0
Re: testing the egs002 inverter board
« Reply #484 on: November 10, 2017, 03:04:37 am »
A nice explanation 'Oztules'.

I will add that to my latest update reprint of the OzInverter Book.

Its funny really the more these OzInverters work the more people keep asking questions about the most intricate internal workings.

'Oztules',    Trying to be as simplistic as possible, ......... Am I correct in saying that the 8010 chip is running at 20KHz, and creating 17 PWM steps per cycle of the 50Hz, to create the Pure Sine Wave by using those big Capacitors as a smoother between steps/pulses.?   

Yea I know there's more going on, but just trying to get a brief, 'man in the street understanding'  . I had a meeting last week with some RE folk here in France.
 Later, I saw the photos of the audience, 50% had that glazed over look, oh dear!

Its not that easy trying to get the message out there.  I do not want to put folk off, especially those who have passion but lack that in depth knowledge.

Offline oztules

  • Moderator
  • Hero Member
  • *****
  • Posts: 1177
  • Karma: +105/-8
  • Village idiot
Re: testing the egs002 inverter board
« Reply #485 on: November 10, 2017, 04:02:28 am »
Short answer is I don't know for sure... never bothered to measure it.

Somewhere I remember in the data sheet I think, that 20khz is about right for the spwm pulses.... but I don;t know for sure..... never interested me, as I can't change it.

The number of steps? don;t have a clue, but I suspect 17 is a bit low. Looking at the cro, it appears more like many hundred... it is the fuzzy either side of the sine wave under some circumstances. My CRO is old and has odd habits, sometimes it shows the fuzz, and sometimes not :o

But if you look, it is there.

The sine wave is shaped by the small 4uf capacitors on the output side.
The 4uf does not sound much, until you see the 20000hz, then Xc becomes more likely to make a substantial  difference to things.... probably in the 2 ohm range.

So the big caps, are only there to  maintain a reliable energy source for the spwm to work with. Not much point trying to increase a pulse width, only to find the voltage greatly sag.... so check feedback....and widen it some more......etc.

So,the big bulk caps do what bulk caps do, and they get some protection from the smaller 4uf on the main board to try to keep the RMS  ripple low in the bulks with the high ESR.

Remember I am not an EE, so I can only say what I see... and now I find that there is no such thing  a gravitational force.... so I can't believe what I see either. .... nor can Newton with his apple........Einstein seems to have canceled gravitational force  a hundred years ago....... I must be the last bloke to find out.




........oztules     
.......... who is being held down by warped space time ???
Flinders Island...... Australia

Offline ClockmanFrance

  • Full Member
  • ***
  • Posts: 171
  • Karma: +11/-0
Re: testing the egs002 inverter board
« Reply #486 on: November 10, 2017, 04:20:02 pm »
Thanks 'Oztules'.

Some where on one of forums someone (with lots of fancy gear) reckoned there were 17 or maybe it was 27, but to be honest I can't flipping remember.  But definitely a 7 at the end.

Good job I scribble/copy/paste things down, but that one slipped by.

Offline ClockmanFrance

  • Full Member
  • ***
  • Posts: 171
  • Karma: +11/-0
Re: testing the egs002 inverter board
« Reply #487 on: November 10, 2017, 04:58:14 pm »
The last Big OzInverter.

75mm/2 primary cable, and 16mm/2  secondary cables coming from 6off (6 in hand) 1.8mm diameter enamelled copper secondaries.

I find that the 6Kw OzInverter really does cope with everything I throw at it, and making them big is just not utilising materials properly, ie, even with all my buildings etc I am just not using all the cababilities of the big 10kW OzInverter. 

Offline oztules

  • Moderator
  • Hero Member
  • *****
  • Posts: 1177
  • Karma: +105/-8
  • Village idiot
Re: testing the egs002 inverter board
« Reply #488 on: November 10, 2017, 07:36:01 pm »
I find that if you can do 3-4kw constant, and run 5-8kw for 5mins or so, it is going to be easily enough to run any off grid house.
I think it is important to have the 5-8kw available for short durations to boil a jug, or run a toaster, while the hotwater is on for a few hours, and the washing machine may be on as well ( heating it's own hot water too).

I don't like the concept of having to make any deviations from what I would have done if the grid were on instead. I don't want to make a compromise of any sort, nor have to school my wife on what not to do... it wouldn't work anyway.

So I find that 16x  hy4008, a twin core tranny, and we seem to have it all under control.


If I have a unit near the cro, I will try to count the steps.


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

Offline ClockmanFrance

  • Full Member
  • ***
  • Posts: 171
  • Karma: +11/-0
Re: testing the egs002 inverter board
« Reply #489 on: November 11, 2017, 03:58:37 am »
Yes + 1.

I had got used to the commercial boys telling us that the 6kW (from that so called Rolls Royce of Inverter manufacturers) was rated at 6kW for a few minutes, and actually it would only really run at 4kW all day if necessary.

My 6kW OzInverter handles all my loads, from my workshop big lathe motors ( max Single phase you can get), underfloor heating in 4 buildings and normal domestic appliances.

With the GTI's AC coupled to my OzInverter Mini Grid, the OzInverter copes well with back charging and the ever fluctuation of loads.

Like you, I am not training my family of 5 to do anything different with domestic appliances.

Sometimes I realise that the OzInverter is coping with 8kW of AC coupling from the GTI's, its not worried about the 5kW of PV that is DC directly charging the batteries, its not worried about the 3off 3.7m diameter Hugh Piggott design wind turbines directly charging the batteries, and the OzInverter doesn't get into a flap when its asked to cope with all of this and give up to 12kW and more.

What's even better the OzInverter does not need expensive and complicated equipment add ons to control the Inverter like the commercial boys.

And yes I have tested it, it can run all day at 6kW.

Nice one Oztules, I do love simplicity, robustness,  and the cost effectiveness, well done matey.

 :-*

Offline Phaedrus

  • Jr. Member
  • **
  • Posts: 28
  • Karma: +0/-0
Re: testing the egs002 inverter board
« Reply #490 on: November 13, 2017, 12:38:24 am »
Fantastic explanation ,Oztules !
Really appreciate the amount of time,you spend sharing your knowledge and coaching newbies.I knew about RMS and Peak Voltage abstractly,but actually applying the concept in this system design is a revelation of sorts.

Oztules,Rossw :
Regarding the CT,I have attached a datasheet.This CT gives an output of 0.05A proportionately.It mentions a burden resistor of 100ohms,which would give 5V ,which is too high for the IFB pin.Is it OK to use a 10ohms resistor to get the required voltage of 0.5v (at IFB pin) for the egs002 board?

Thanks very much indeed for the patient support!

Thomas

Offline rossw

  • Senior Moderator
  • Hero Member
  • *******
  • Posts: 879
  • Karma: +35/-0
  • Grumpy-old-Unix-Admin
Re: testing the egs002 inverter board
« Reply #491 on: November 13, 2017, 03:03:58 am »
Oztules,Rossw :
Regarding the CT,I have attached a datasheet.This CT gives an output of 0.05A proportionately.It mentions a burden resistor of 100ohms,which would give 5V ,which is too high for the IFB pin.Is it OK to use a 10ohms resistor to get the required voltage of 0.5v (at IFB pin) for the egs002 board?

If it will produce 5V into 100R (requiring 50mA to do so), then replacing the 100R with 10R will indeed produce 0.5V at the same 50mA, so yes!

Offline oztules

  • Moderator
  • Hero Member
  • *****
  • Posts: 1177
  • Karma: +105/-8
  • Village idiot
Re: testing the egs002 inverter board
« Reply #492 on: November 13, 2017, 04:05:23 am »
If your using my circuit, here:

7279-0

You will find any CT coil will do. It is designed for a very wide range.

The IFB does NOT see the  output of the CT. only the SCR sees the  output. Have a good look at the diodes, their orientation, and their circuit path.

The CT complex acts as a floating power supply, ONLY the scr sees the potential from the rectified outputs. The IFB sees the neg CT, but it has no ground reference, so is meaningless to it.

Any CT will do, the SCR will trigger and look after the rest of it.

When triggered, the CT is shorted anyway. I thought the floating supply was pretty nice actually, as it allowed me to do a few things easier..


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

Offline Phaedrus

  • Jr. Member
  • **
  • Posts: 28
  • Karma: +0/-0
Re: testing the egs002 inverter board
« Reply #493 on: November 14, 2017, 09:54:48 am »
Hi Oztules,
I was actually referring to the schematic which was posted on the backshed forum,attached here.
The CT output is rectified and sent to the IFB pin as can be seen.I dont think I have seen the PCB revision which has an SCR for triggering the IFB.Could you post a complete image with a BOM,when convenient.There are so many version,its easy to get lost :).
Does the IFB trigger a soft start too?Functionally is there a difference between an over temperature trip or a current trip?
Thanks!



Offline oztules

  • Moderator
  • Hero Member
  • *****
  • Posts: 1177
  • Karma: +105/-8
  • Village idiot
Re: testing the egs002 inverter board
« Reply #494 on: November 14, 2017, 03:11:44 pm »
If your using torroids , I would not use that circuit. It is too slow, and the shut down causes false triggering I have found. This does not worry a normal transformer setup, but a torroid presented with any wave not consistent with the previous waves, will blow everything up.

It should work fine, but it did not. It worked well some times, and blew it up other times... so it was ditched.

I moved to a system where the unit will shut down very very fast, and be latched down, and would stop all signals going to the fets, by shutting down the 2110 drivers, and simultaneously the ifb of the 8010.... it shuts down and stays down. It will not restart until all power is drained from the control board.

Over temp is soft start, and retries consistently. IFB is soft start with about an 8 second restart timer incorporated. It should be fine, but has blown a lot of fets testing it out.

Attatched is the manufacturing gerbers used for making the boards as I currently run them.
Use the image I posted  a few posts back to correct the few component changes.

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

I need to do a proper how too, as your right there are many many versions as I progressed out there. It needs to be in one place as a definative vresion. Both of the solar controller for it, and the inverter as a set.


Flinders Island...... Australia