Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.
tinyt Guru Joined: 12/11/2017 Location: United StatesPosts: 438
Posted: 04:07pm 22 Sep 2018
Copy link to clipboard
Print this post
Thanks Poida for the useful info.
on a side note, I am planning to modify the code you shared and adapt it to an atmel chip (not used in arduino, unfortunately). Hopefully, it will emulate most of the 8010 functions but with the advantage of not dealing with a 'black box' but knowing the hardware internals and software source code, that can be corrected, improved, etc.
It might take a while as I am just a C beginner and I also need to finish first the inverter I started over 6 months ago.Edited by tinyt 2018-09-24
Clockmanfr Guru Joined: 23/10/2015 Location: FrancePosts: 429
Posted: 07:33pm 22 Sep 2018
Copy link to clipboard
Print this post
Hi Wiseguy,
The PM inbox is now sorted, thanks for letting me know.
The ZIP file is all GERBER RS-274X format files ready for the manufacturer. Normal Gerber viewers can see them.
My PCB program is the German TARGET 3001 version 18, I like it as it interlinks with my Engineering CAD. If you can read TARGET I will put the file up?
Edited by Clockmanfr 2018-09-24Everything is possible, just give me time.
3 HughP's 3.7m Wind T's (14 years). 5kW PV on 3 Trackers, (10 yrs). 21kW PV AC coupled SH GTI's. OzInverter created Grid. 1300ah 48v.
poida Guru Joined: 02/02/2017 Location: AustraliaPosts: 1418
Posted: 10:04pm 22 Sep 2018
Copy link to clipboard
Print this post
What am I testing for? To see if the EGS002 output (from EG8010->transistors->IR2110->mosfet gates) i.e. pins 3,6,8,10 contain no cross conduction and if so does it vary as dead time varies?
or
how dead time varies at pins 3,6,8,10 as dead time is set via the jumpers.
I am not sure what you want to see.
I am confident no cross conduction drive conditions are output on these pins of the EGS002 no matter the dead time setting.wronger than a phone book full of wrong phone numbers
wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156
Posted: 12:11am 23 Sep 2018
Copy link to clipboard
Print this post
Hi Poida, perhaps I can answer for Mark. I think he wants to know in a setup like his without the cross conduction transistors fitted, how much of the 180nSec pulse is seen by his FETs. I believe you have a similar unit at your disposal you could investigate with ? If you are going to do this, the results for before and after the 5R6 gate resistors would help complete the picture.
Lastly I have only spent a few minutes after I discovered the overlap issue with the EG8010. What if other failure, hiccup or power up/down events could cause a longer overlap ? - I need to get back to the bench too.
Gaspo, I have TL431's, I have ordered 12Mhz xtals (HC49S) but sometimes parts from China arrive in a week others take ages. I dont have any 1206 parts on hand at this time.
The TL431 resistors could use the 2 x 5K1's - they waste ~ 400uA. The suggested series resistor 3K3 to pin 17 & a 12V supply, passes ~ 2mA in total. Output should be around 5.01V using typical parameters. Edited by wiseguy 2018-09-24If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike
renewableMark Guru Joined: 09/12/2017 Location: AustraliaPosts: 1678
Posted: 03:19am 23 Sep 2018
Copy link to clipboard
Print this post
The whole reason for this thread is because Wiseguy is afraid of this happening.
My question is that actually a problem at the fets? Are they turning on at all? If they are how much and does the dead time setting change it?
BUT you need a running Madinverter to test it. The fact that people (not just me) that run the machine have no problem with these restarts makes me think it's either a small problem or not one at all.
I'll leave you all alone now.Cheers Caveman Mark Off grid eastern Melb
wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156
Posted: 05:48am 23 Sep 2018
Copy link to clipboard
Print this post
The whole reason for this thread is that any engineer worth his salt, will strive to avoid shoot-through like the plague, there should be no discussion about well its not a large amount of overlap.
The Deadtime setting has no effect on this problem because it doesnt know about the bug that exists
Think of the logic - Klaus has set his dead-time - lets define dead-time on the way - its adding safety time to avoid overlap or shoot through - to 500nSecs to be ultra safe. Yet he has an overlap occurring for 180 nano seconds before he even starts adding the 500nSec protection time - and you argue this is a good idea and we should leave it alone.
I will leave it to the Forum to draw their own conclusions and whether they are happy to have this condition remain in their inverter I'm not fussed either way.
I stated from the outset that part of the reason for posting the information was to help others designing with the EG8010.Edited by wiseguy 2018-09-24If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike
zaphod Regular Member Joined: 03/06/2018 Location: United KingdomPosts: 93
Posted: 06:46am 23 Sep 2018
Copy link to clipboard
Print this post
Unfortunatly wiseguy places like this are populated by non-engineers who like engineering managers are sceptical about what they are told by professionals with no basis, they are only interested in saving money or having you spend your time on some pet project of there's, company politics all over again!! Keep up the good work and leave the luddites with there heads in the sand :)Cheers Roger 1Kwp DIY PV + Woodburner + Rainwater scavanger :)
renewableMark Guru Joined: 09/12/2017 Location: AustraliaPosts: 1678
Posted: 08:44am 23 Sep 2018
Copy link to clipboard
Print this post
I said I was going to leave you all alone here, But I feel compelled to reply to this.
"No basis" well factual evidence of repeated restarts is in fact evidence, clear in the field evidence, not some theory in a book.
Luddites??? really??? I would call myself the exact opposite, I have embraced this technology learned it and made my house completely off grid 8 months after picking up a soldering iron, maybe learn what a word means before you toss it willy nilly at someone. Also note I have not used name calling to anyone here. I have asked why my machine or the dozens of other ones out there works so well with this apparent problem, but no one really has given me a reason. No one has tested if this phantom menace actually exists at the fet because none of you have built this, all just theory..... If anyone has an improvement to this design, fantastic! But wouldn't it be wise to see if the current circuit boards actually have that problem? I would have thought that was a pretty reasonable question. If there is a problem, then yes deal with it.Cheers Caveman Mark Off grid eastern Melb
Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498
Posted: 08:58am 23 Sep 2018
Copy link to clipboard
Print this post
zaphod are you on a mission to insult a large number of people on this forum with name calling? If you feel the DIY nature of the forum is beneath you then why are you here? From what I have seen of your inverter attempts you have been far from successful yet you feel qualified to attempt to insult others that have built inverters that do work.Edited by Madness 2018-09-24There are only 10 types of people in the world: those who understand binary, and those who don't.
Tinker Guru Joined: 07/11/2007 Location: AustraliaPosts: 1904
Posted: 10:22am 23 Sep 2018
Copy link to clipboard
Print this post
Perhaps you might want to consider, gaspo, that there are quite a few operating inverters about and for these a plug in shoot through protection might be considered.
But, that would not require a crystal, since the control board already has one, and any changes to the function of any of the 32 pins would involve cutting control board tracks. So, as nice as the TL431 idea sounds, it immediately renders the little board no longer a plug in EG8010 replacement.
I would be happy to experiment with a *plug in* replacement (EG8010 plus shoot through protection) but not if any control board tracks need to be modified.
Maybe two different versions of that little board?Klaus
wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156
Posted: 11:15am 23 Sep 2018
Copy link to clipboard
Print this post
Zaphod I'm afraid that I must agree with Mad's and Mark's last two posts.
I do share your frustration though at this being hard work at times trying to explain technical engineering issues to non engineers. But we still need to be civil even if at times we are frustrated. Things like that are better muttered under your breath or in a PM.
If you read enough postings here you will see that Mad, Mark & Tinker and many others (even if some have been on my case a lot) are extremely talented and helpful individuals with multiple skills & knowledge and have carried this project over a number of years.
Klaus the board that Gaspo has created can be used without cutting tracks etc.
If you get a blank board and load your own parts, by not loading the TL431 and putting a link or zero ohm resistor between the pin 17s (R11?) it works as before. If you dont want to put the crystal & its 2 capacitors on the adaptor board & take advantage of the lower noise pickup you can jumper the 2 crystal connections with 2 short links to the unconnected pins 10 & 11. For the best rejection of noise though having the crystal and capacitors on the adapter board is a good idea. The crystal and caps on the main board would just become unused - you dont have to remove them. If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike
zaphod Regular Member Joined: 03/06/2018 Location: United KingdomPosts: 93
Posted: 11:31am 23 Sep 2018
Copy link to clipboard
Print this post
I think my wording could have been better, I meant to observe the way I felt the wind blowing on this forum rather than intentionally insult anybody, however I don't wish to hijack this thread other than to agree what good work it is and feel astonished that others seem to want to pour cold water upon it (but maybe that is my faulty interpretation of posts). Yes that particular GTI has had problems but please remember it is my own design not a copy of somebody else's work so it is bound to have some teething troubles particuarly as I have no previous experience of direct grid coupling as a generator. No I don't feel this forum is beneath me and anyway I used to be a digital designer in computer communications very little to do with power electronics. I am however very interested in DIY greeness of whatever flavour. Best wishes to all :)Cheers Roger 1Kwp DIY PV + Woodburner + Rainwater scavanger :)
gaspo Regular Member Joined: 25/06/2018 Location: AustraliaPosts: 65
Posted: 12:50pm 23 Sep 2018
Copy link to clipboard
Print this post
Tinker, I should be able to connect tracks from the crystal pins on the 8010 to the corresponding connector pins. That way you wouldn't need to modify your control board at all if you don't want to.
Thanks all for the constructive suggestions. The adapter board seems to evolved to quite universal form catering for both the new designs as well as drop in replacement for the existing inverters.
Tinker Guru Joined: 07/11/2007 Location: AustraliaPosts: 1904
Posted: 01:14pm 23 Sep 2018
Copy link to clipboard
Print this post
Thanks for the quick reply gaspo. Did I read somewhere that you are located in Perth WA? If so let me know when you have such an adapter board ready and I can bring my little inverter to test it on - assuming you do not have such yourself.
BTW, on my control board the crystal is under the EG8010 adapter and as close as possible to the pins 10&11 so having the crystal on the adapter itself would make no real difference IMO.Klaus
gaspo Regular Member Joined: 25/06/2018 Location: AustraliaPosts: 65
Posted: 01:53pm 23 Sep 2018
Copy link to clipboard
Print this post
Tinker, yes I'm in Perth, WA. I'll let you know when the adapters are available for testing. I've sent you PM with more info.
Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406
Posted: 08:44pm 23 Sep 2018
Copy link to clipboard
Print this post
We know there are different versions of the EGS002 boards, some have these anti shoot through transistors fitted and some do not. There are also some missing components on most if not all of these boards.
I am wondering if there might be more than one software version as well. All these changes do suggest attempts at fixing known bugs, or maybe just improving the product.
There could also be counterfeit EG8010 chips out there, the Chinese are notorious for copying and any counterfeit chip may have totally different software in it that almost does the same thing as the genuine original but not quite.
We may all be arguing with each other on a false premise here. It may be that only some of these chips have the shoot through problem.Cheers, Tony.
wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156
Posted: 12:12am 24 Sep 2018
Copy link to clipboard
Print this post
Not trying to be argumentive Gaspo but did I miss something ? As it stood with pins 10 & 11 unused, if the crystal is fitted & used on the adaptor board there is no modification to the control board. If the adaptor board has no crystal or capacitors fitted & pins 10 & 11 are linked there is still no modification to the control board ? What modifications were you referring to that could otherwise be necessary for the control board ?
I applaud Tinker for wanting to keep his crystal circuit on the control board but we have a cross section of 1 comment that I am aware of and we are going to now put longer antennas to pick up noise on our previously improved xtal layout that affects others who may not want this feature. I am sure Tinker could put 2 small links to pins 10 & 11 - he may be the only one to see a valid reason for saving a dollar to do this ?
You may have made another valid point Warp - its still a work in progress.Edited by wiseguy 2018-09-25If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike
wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156
Posted: 01:26am 24 Sep 2018
Copy link to clipboard
Print this post
Warp I was just about to move my attention to output stage response of the glitch. Although the TIP41/42 transistors are a bit agricultural and I would never consider them for gate drivers (stepper motor bridge maybe....) I opened the data sheet.
FYI the data sheet for a TIP41/42, it has an Ft of 3MHz. My understanding is that the FT is the frequency that HFE drops to 1. 3MHz also has a period of 300nSecs, there is also the inherent delay and rise times to consider. So maybe we have more insight into how inverters using these slow transistors are surviving.
If you were to use a fast FET driver such as the TC4452 with a rise time of 20nSecs into 10nF & 40nSecs into 22nF I am confident FET's would not survive. - I believe in some posts here TC4452's were tried but it killed output FET's and the drivers perhaps the glitch discovery now explains why that happened.
Do two wrongs make a right? Not in my book, but we have a glitch and we have a slow gate drive that appears to respond so slowly in FET time who knows what the actual final result is re shoot through overcurrent. For every TIP41/42 the result will be slightly different with varying values of Ft, gain etc. I was going to breadboard this to analyse performance but with so many varying parameters, including different layouts wire lengths etc, I think the time spent and the results would be inconclusive. I think my time is better spent developing my own version of an inverter, where I will actually get something tangible for my efforts.
What are your comments ? Edited by wiseguy 2018-09-25If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike
Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406
Posted: 03:20am 24 Sep 2018
Copy link to clipboard
Print this post
Agree with all of that 1,000%
I don't really have a dog in this fight, but all of my "large geometry mosfet" designs always use something like a TC4452. If you go back and look at the picture of my wimpy little 1.5Kw PWM inverter back on page four of this thread, there are four TC4452 chips in plain view.
The only exception to that is when I am switching at only perhaps 50Hz (not 20Khz+), where a very slow turn on and turn off, plus a couple of microseconds of dead time is no problem at all.
Edited by Warpspeed 2018-09-25Cheers, Tony.
wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156
Posted: 03:51am 24 Sep 2018
Copy link to clipboard
Print this post
Thanks Warp - sometimes I feel the need for a sanity check. I consider that whatever the answer is as to what happens down the chain after the Glitch occurs in existing inverters is not part of my problem anymore, so I can ignore it.
I will be using TC4452's driving at least 4 FETs each. I dont regret finding or announcing this - had I not found this, I know I would have been bitten with the glitch in a spectacular way, and I may have helped anyone wanting to do their own version of hardware.
I'm also not finished with my evaluation of the EG8010, so its back to the bench this afternoon and a few less postings from me for a while.
If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike
Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.
Print this post
