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.
Forum Index : Electronics : Whispergen Stirling gen. coupling to 48vdc with 50 hz H bridge in usa
Page 1 of 2
Author
Message
joeblack5 Newbie Joined: 08/02/2022 Location: United StatesPosts: 19
Posted: 12:37pm 08 Feb 2022
Copy link to clipboard
Print this post
Hello folks , New member here. Introduction.
Retired electronic engineer but more fun with mechanics. 10kw of solar, outback gridtie inverters with 48 volt battery bank as backup. Natural gas water-cooled Onan generator set up in Cogen charging the 48 volt battery bank during night and delivering back thru the grid thru the outback inverters. Additional some vacuum tube solar collectors for domestic hot water heating.
In my next post I will describe my goal and problem and request for help with the Stirling generator.
Johan
joeblack5 Newbie Joined: 08/02/2022 Location: United StatesPosts: 19
Posted: 12:09am 09 Feb 2022
Copy link to clipboard
Print this post
Lost my second post, got logged out, live and learn.
Always have been fascinated with stirling engines and also cogen. Build and played around with the toy stirling engines.
Several years ago I bought in Germany two used and broken whipergen cogen generators. These are normally gridtie with 50 Hz and 220Volt and run on natural gas. Whispergen is not in business anymore but a couple of dedicated individuals try to keep remaining units serviced.
Retired now and a bit more time available I like to get the whispergen running. In my simplistic thought I had hoped that the Grid tie electronics would have setting from 50 to 60 HZ. After talking with some service people it turned out that there is no 50 / 60 hz setting and that the generator part inside the stirling engine is synchronized with the grid. If it would be poible to run the tirling from 1500=50Hz up to 1800=60hz then it would get less efficient because of increased pumping losses.
So now i hope to create a 50hZ microgrid to satisfy the whispergen grid tie requirements and start the stirling motor then have the 50 hz power backfeed thru the H bridge into the 48 volt battery bank.
After reading thru a lot of threads about egs002 inverter boards and their problems I decided to buy a 48 Volt 24 mosfet board with the totem pole driver.
The output of the whispergen is maxed out at 1000 watt. so I assume to start it would be comparable to a 2 HP compressor.
I will disable the Ifb as described although I do have some questions about that.(later)
Are 15 Volt zeners still useful to add to the board when it has the totem pole drivers?
Anybody any experience with whispergen in our outside what i try to do please chime in.
Thanks Johan
Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406
Posted: 12:38am 09 Feb 2022
Copy link to clipboard
Print this post
I depends where you are I suppose, but your stirling engine is likely only going to be useful in a clear blue sky. More solar panels might be a more practical and less troublesome alternative.
Never heard of a whispergen, but running a gasoline engine on low cost natural gas is definitely a good idea here in Melbourne.Cheers, Tony.
joeblack5 Newbie Joined: 08/02/2022 Location: United StatesPosts: 19
Posted: 01:15am 09 Feb 2022
Copy link to clipboard
Print this post
Hi Tony,
thanks for chiming in, and thanks for the many many contributions you have made to the EGS002 and your own very ingenious inverter design.
We live in State College Pennylvania with 3 month of winter where the solar panels do not deliver enough power for our solectria electric vehicle, heatpump or induction stove. I have a Cogen Lister clone 6Hp liquid cooled on diesel / wvo and recently started an Onan 6KW cogen liquid cooled on natural gas. In the summer the solar can hardly keep up and so we are expanding out solar to 15Kw and possible more.
The stirling engine is more a hobby but possibly can be a participating during the winter time.
The asynchronous generator in the whispergen is 4 pole single phase. easy to google whispergen stirling if interested. designed in New Zealand.
Johan
Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406
Posted: 01:42am 09 Feb 2022
Copy link to clipboard
Print this post
Hi Johan, one way to overcome the 50Hz/60Hz problem is to adopt the inverter alternator principle.
Rectify whatever comes out of the wispergen alternator into a high voltage dc bus, and then either run a pwm sine wave inverter off that to generate your 120v 60Hz, or a dc/dc buck regulator for battery charging.
All that machinery is really going to shock your Amish neighbors. Edited 2022-02-09 11:44 by WarpspeedCheers, Tony.
joeblack5 Newbie Joined: 08/02/2022 Location: United StatesPosts: 19
Posted: 03:14am 09 Feb 2022
Copy link to clipboard
Print this post
Haha, yes it does, the shocking part..pun intended.
The problem is these unit use the grid to get started. Very similar to solar grid tie.. no grid no power.
So besides making it hot I have to give it an electric kick to start it. The generator part doubles as the starter motor. Of course for the controller to even come on it wants to see a nice 50 Hz sinewave. Just as gridtie the machine will stop if you disconnect it from the grid. Johan
joeblack5 Newbie Joined: 08/02/2022 Location: United StatesPosts: 19
Posted: 03:22am 09 Feb 2022
Copy link to clipboard
Print this post
I thought about the high voltage DC bus by taking a vfd and setting it on a fixed frequency of 50 Hz , maybe running a synchronous motor with flywheel in parallel and connect it to the whispergen. When that would start up it would deliver back to the high voltage bus and from there convert back to the 60 hz grid.
More or less the same as I do now only instead of a high voltage bus I use a 48 volt bus and use the outback gridtie inverters to deliver back to the grid.
Somewhere in the egs002 discussion there is talk about allowing the frequency to shift a little and running a bldc motor with flywheel in parallel with an inverter to create more startup / surge power for induction motors. Not sure what thread and I do not think anybody picked up on that idea.
Johan Edited 2022-02-09 13:23 by joeblack5
Godoh Guru Joined: 26/09/2020 Location: AustraliaPosts: 458
Posted: 05:41am 09 Feb 2022
Copy link to clipboard
Print this post
Hello Johan, I have found a way to get more startup current to my inverter. I have a 500 farad capacitor bank on parallel with my 24 volt battery bank. It works a treat. The capacitors supply a massive surge when needed. Up to 1900 amps according to the specs. 48 volt banks of these caps are used to start trains, I believe the ones we get in Oz come from the US. So they should be pretty easy to source there. I think that they are Maxwell branded capacitors. Pete
noneyabussiness Guru Joined: 31/07/2017 Location: AustraliaPosts: 512
Posted: 07:17am 09 Feb 2022
Copy link to clipboard
Print this post
I remember there was discussion about that, but if you really hard up ask " poida " he has a arduino controlled inverter that he could maybe add that to the programming... or as godoh said, a bank of really big caps would probably work...I think it works !!
joeblack5 Newbie Joined: 08/02/2022 Location: United StatesPosts: 19
Posted: 01:02pm 09 Feb 2022
Copy link to clipboard
Print this post
Thank you for these thoughts. I will have to look for the train capacitor bank and see what floats around.
I will have to read the problems about ifb in the egs002 again.
Sofar I got out of oztules that the 393 turned both the 2110 driver chips and eg8010 and because of bouncing inverter restart that specifically toroid had issues because of the tight coupling and stored magnetic energy.
Then poida or wiseguy found that the eg8010 had timing issues when ifb was activated resulting in cross conducting moments..bad of course.
Have to drop the kids at school and don't want to loose my scribbles.
joeblack5 Newbie Joined: 08/02/2022 Location: United StatesPosts: 19
Posted: 02:22pm 09 Feb 2022
Copy link to clipboard
Print this post
Back from the school run. I think wiseguy then came up with a four transistor schematic that prevented cross conduction of the eg8010 chip...funny enough that seems to be integrated in the egsO02 board but was absent in the EGS001.
So here we have that oztules experimented with the egs002 board and discovered problems with the ifb ..but that cannot have been an issue with cross conduction because the four transistor part already took care of that. Then only the quick restarts could have caused the havock with tight transformers like toroids.
It is unfortunate that the thread with wiseguys 4 transistor solution and then another set of transistor to make it perfectly symmetric ended suddenly without test and resolution.
You all probably say ..why are you talking about stuff from 4 years ago.
Well ..my only goal is to start an inductive motor load that is connected to a computer control board. I really like to get a good understanding. Blowing up a couple of MOSFETs is ok but I rather not blow up the whispergen control at the same time.
Johan
Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406
Posted: 06:53pm 09 Feb 2022
Copy link to clipboard
Print this post
Those bamboo EGS boards of any flavor seem to be cursed with a number of different issues, and I think many people here have now lost faith in the things.
I have not been following the ongoing PWM saga too closely, but I believe there are now some good Australian Nano based driver boards and software around that should provide a much more solid basis to start from.
Also, the magnetics design of both toroids and chokes has come a very long way since the early Oz days. Idling power has been substantially reduced, and inverter robustness increased.
Any choke is better than no choke, but a really good NON SATURATING choke can really help the mosfets survive really heavy surge loads.Cheers, Tony.
joeblack5 Newbie Joined: 08/02/2022 Location: United StatesPosts: 19
Posted: 08:38pm 09 Feb 2022
Copy link to clipboard
Print this post
Thanks warpspeed. Just checked in my shop and I have some pretty substantial chokes around. I wound a secondary on a transformer when I was 18 yr for a power amplifier, never used it. Now 44 yr later I will give it a second chance. I read about the nano boards and will see how much where and so on and how compatible. It seems that most people just disabled the ifb on the 8010 and live with the main breaker. This transformer is long ago. The core is 1-1/2 by 3". Vaguely remember 1.5kva. Will measure the idle current thru the primary. Edited 2022-02-10 06:39 by joeblack5
Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406
Posted: 08:50pm 09 Feb 2022
Copy link to clipboard
Print this post
Check out some of the Forum threads on Nano driver boards. I am a Warpverter heretic, and don't follow the PWM/Nano threads with very much interest.
Also plenty of old threads on design and fabrication of toroids and chokes which is more my area. Its like digging for gold, its definitely down there, but it will take some finding.Cheers, Tony.
joeblack5 Newbie Joined: 08/02/2022 Location: United StatesPosts: 19
Posted: 10:03pm 09 Feb 2022
Copy link to clipboard
Print this post
I think I better stay with the EI core. Safer that the transformer coupling is less stiff for this application.
The EI core I had in mind is only 3 x1-1/2". Barely 600 watts at 50 Hz.. have to look at something bigger.
Hope that poida and wiseguy drop in what the final story was with the ifb.
How does one upload pictures in this forum?
Johan
Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406
Posted: 10:14pm 09 Feb 2022
Copy link to clipboard
Print this post
EI cores will probably Hummm, and idling power will be higher. But it will obviously be a lot easier to wind.Cheers, Tony.
poida Guru Joined: 02/02/2017 Location: AustraliaPosts: 1418
Posted: 11:19pm 09 Feb 2022
Copy link to clipboard
Print this post
Hi joeblack5
we are in a position where we can make an inverter do whatever we need. We could make a custom version that is 60 Hz or whatever you need 240V output with OR WITHOUT voltage feedback (no Vfb might be best for backfeeding) sync the inverter to a master mains freq source
re. ifb
my views are built from my experience of building and blowing up my house inverter as well as extensive testing on the workbench. I am not an EE.
The first few inverters I built used the EG002 module that has IR2110 gate drive ICs for both half bridges. The inverters did not have totem pole gate drive on the boards. Gate drive was direct from the IR2110 output pin, to the FET, with a 4R7 resistor plus small diode.
IR specs for the 2110 show strict limits on voltages on the gate drive outputs. max limits are < Vcc + 0.3V and > -0.3V Internally it's likely a FET totem pole output stage. I have found evidence that FET Gate voltages exceed specs, more often going much more negative than the 0.3V limits I also have seen FET Gate voltages exceed Vcc+0.3 as well. Over time, these voltage excursions destroy the oxide Gate insulation. It might take 1 second or 1 month. The voltages originate from the parasitic capacitors within the FETs and as you drive the inverter harder, more current is switched and so higher voltages are sent OUT of the FET Gates into the IR2110 output pin's driver stage. The fail short and that means the inverter's FETS are driven both ON. And it's time to build another inverter...
I protected against this as my first approach to use TVS diodes. They are fast and have lowish Vforward which is good. I put diodes across all 4 output pins of the 2 IR2110 in such a way to prevent negative voltages. The diode Vf was about 0.5V but maybe smaller with the small currents involved here. Dunno. But when I did this, the inverters lived. And lived much longer.
It became clear that a home built power board (Madness's power board to be precise) was the way to go. 1 - cheap to build 2 - repairable after a blowup 3 - on-board robust totem pole gate drive made with discrete parts and so 4 - easy to debug
As soon as I built my first Madness board, I probed the Gate drive signals at the IR2110 pins No more damaging negative voltages. They for sure are still produced "in-FET" but they can not pass out the totem pole drive and then via the two transistor Base pins to the gate drive ICs This is in my view the reason why the on-board totem pole drive power boards are so reliable.
but what about ifb?
Experience has shown to me that the HY4008 FET is quite strong when driven properly. I built a 4 HY4008 by 4 board (16 FETS in total) and tested it to see what will blow first. A 50V DC 240V 50Hz system, with a 2kW toroid. A 120A DC breaker on the input and a 20A 240V AC breakeron the output. It started the single phase air compressor with ease. I recorded input current and saw 12kW peaks during motor startup. And it settled to 1kW during running. From these tests I see no reason to protect from over current apart from a simple "break input supply upon power board failure" type of thing. So that is what the 120A breaker is for.
The EG002 has the function to limit inverter output power based on a current limit. It works in 2 ways. Once ifb exceeds 0.5V PWM output is disabled. It's disabled immediately via the shutdown pins of the IR2110, and it may happen mid PWM pulse. It's a fast acting shutdown. Also, the EG8010 takes this signal and shuts down PWM at the end of the 1/2 cycle (I could be wrong with that) but the PWM output from the 8010 IS stopped by the chips firmware sometime or other.
I found in testing that ifb could be a noisy signal, or have fast changes. This implies PWM gate drive could be quickly interrupted and reinstated. I once had a 1.5kW toroid seem to jump off the workbench when testing ifb settings. I decided to disable it and use the DC supply breaker No more ridiculous full power fast rising power transients, ever.
so that's a long intro to my views on the EG002 and why I made the nanoverter project. Have a look at the picoverter too, far less parts and just as reliable. Edited 2022-02-10 09:23 by poidawronger than a phone book full of wrong phone numbers
joeblack5 Newbie Joined: 08/02/2022 Location: United StatesPosts: 19
Posted: 12:32am 10 Feb 2022
Copy link to clipboard
Print this post
Thank you Poida for chiming in and summarizing this all again.
I previous read all the experienced people posts / threads about these EGS002 boards. Based on your observation I bought the 4x6 Mosfet unit with the external totem pole TIP41/42 driver. Is there still a benefit on adding TVS diodes to an external totem pole drive?
I do not have the actual board yet , in the mail, Chinese new year .
I will do as you recommend in your post and as in your previous threads and disconnect the ifb and use a 120 Amp DC breaker on the 48V side.
Sorry my programming knowledge is very very limited. Will read and try to get a grasp of the nano and pico inverter.
Is uploading an image as simple as copy and pasting it in the post itself?
Johan
Murphy's friend Guru Joined: 04/10/2019 Location: AustraliaPosts: 648
Posted: 02:19am 10 Feb 2022
Copy link to clipboard
Print this post
Johan, on top of the window you type your text in you'll see a row of icons. click the second last one (green tree with blue arrow) and follow the instructions 'precisely'. It often takes a few seconds for uploading to finish, scroll down your page to see the code for your pic. It will show as a PICTURE after you saved your post.
poida Guru Joined: 02/02/2017 Location: AustraliaPosts: 1418
Posted: 02:27am 10 Feb 2022
Copy link to clipboard
Print this post
I would not bother with the TVS diodes on that board. No need since the TIP41/42 transistors are between the FET voltage sources and the EG002 gate drive outputs.wronger than a phone book full of wrong phone numbers
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
