Puppeteer

Author Topic: DIY Low EMF 5 step Sine Inverter  (Read 10050 times)

ajaffa1

  • Hero Member
  • *****
  • Posts: 1725
    • View Profile
Re: DIY Low EMF 5 step Sine Inverter
« Reply #15 on: August 15, 2018, 10:45:01 AM »
Hey BruceM, sorry to here you have had such a rough time with your mental/physical health and the Medical professionals that are supposed to help.

I have had similar experiences and wonder at the way pills are randomly trialed on the sick. My doctor tried to prescribed a drug called lyrica, for nerve pain. I declined his kind offer as I knew that this product can cause heart attacks and I`ve already had one. He then put me on an antidepressant called Sertraline which put me back in hospital with blood pressure issues.

I went to see the cardiologist today for a check up, told him what had happened and he was appalled that my GP had tried these medications on me without consulting him.

There are some very smart doctors out there who can help, the difficulty is finding them and then having the courage to tell the mediocre medics to take a running jump.

The best news today is that I have been taken off the blood thinners. I will no longer bleed for a week from an insect bite or develop a haematoma instead of a bruise. Happy days  :)

Keep up the good work, chin up,

Bob

BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: DIY Low EMF 5 step Sine Inverter
« Reply #16 on: August 15, 2018, 04:16:46 PM »
I think good medical practice is an extreme exception, rather than the norm. I've seen so much blatant, basic malpractice it's outrageous.  Trusting any medical practitioner is a fools paradise, in my book.  It's your health, and MUCH can be done to help yourself. 

I'm keeping an open mind on the cause of the AC ripple on the DC.  But I know it was NOT there in my bench testing; I had it down to 1-2 millivolts peak to peak at 1000 watts at the end of about 300 feet of cable.  I was expecting an unmeasureable situation with the inverter near the battery bank.  But given the constant level of AC ripple regardless of load, it is almost certainly capacitively coupling from the constant 230VAC pair in the same metal conduit.  I was foolish to have not run a separate conduit for AC.





BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: DIY Low EMF 5 step Sine Inverter
« Reply #17 on: August 15, 2018, 09:17:38 PM »
Not fully believing my own findings (I was still recovering from a trip to town.) I retested today and find results similar to the bench testing. AC ripple on 120vdc supply of about 3 mv p-p at 1000w load, 1mv with 500w.  I suspect operator error or a bad test lead before.

EMI on DC is non-existant,  AC is barely there with AM radio oriented just right a few inches from the wire and tuned just right to the lowest portion of the AM band under even small load.

The EMI is around 1MHz, so as one should expect, it didn't significantly jump to the DC wires in the same conduit, or what did is the same as what bled off to the buried metal conduit.

This is substantially better than my partially EMI suppressed Listeroid/ST-3 power.

I have yet to add the 0.1 ohm capacitors to ground on the H bridge output filter boards, which I will try, since the earth grounding system was such a project! Right now they are just a toroidal common mode choke and snubber.  The only ground connection at present is the center connection of the two transformer secondaries. 


ajaffa1

  • Hero Member
  • *****
  • Posts: 1725
    • View Profile
Re: DIY Low EMF 5 step Sine Inverter
« Reply #18 on: August 16, 2018, 12:19:09 AM »
Hi BruceM, used to get the same sort of problems with control cables in factories, always ran unshielded twisted pairs at least a meter away from single or three phase cables. The worst offender was florescent lighting with starter modules.

Where we had no choice but to lay cabling next to power cables, we ran shielded cables all grounded to earth. Gets real expensive in some situations where you might need 30 or more control/data wires.

Bob

BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: DIY Low EMF 5 step Sine Inverter
« Reply #19 on: August 16, 2018, 04:25:13 AM »
Hi Bob, Yep, industrial control is often one of the worst EMI environments out there; fiber, both plastic and mulitmode glass, has made big inroads there for good reason.  I use Industrial Fiber Optics POF products for both analog/control and digital data regularly. My home's audio, thermostat, central vac, computer on/IR controls, keyboard/trackball and IR rear projection controls are all IFO product based, my own designs.  In my case it's because the home is shielded to -110dB so all signals have to either be filtered or fiber. 

Today I did add 0.01 uF caps to the earthing system on the AC output of the H-bridges.  Connection was easy as I ran sheet copper (16 oz/SF copper flashing) up behind the steel box and could just screw to the box and copper anywhere. It did improve both the DC (nothing audible at all now) and AC.  AC with any load is only audible on the AM radio around 560KHz, within 1/2 inch of the wire with the radio oriented just right.  No load EMI on the 230VAC can be picked up at 6 inches.  I'll have to think about that one.  During testing I experimented with a snubber on the AC output, a motor run cap about 6 uF and 2-10 ohms on my earlier 120V version.  It worked well but I don't like that solution as it's a power sucker.

I thought about adding the 4 - 250 uH toroidal chokes to the 2 h-bridge output filter boards again.  The chokes "round the corners" of the stepped waveform quite nicely but do nothing for the EMI whatsoever.  They do get hot at full load and add about 1% power loss. While the waveform is softer, it doesn't do anything appreciable for THD or motor efficiency so I'm reluctant to add them in. 













ajaffa1

  • Hero Member
  • *****
  • Posts: 1725
    • View Profile
Re: DIY Low EMF 5 step Sine Inverter
« Reply #20 on: August 16, 2018, 12:39:58 PM »
Hi BruceM, I am struggling to understand your set up. I know that your health issues make you very susceptible to EMI. Why do you have any AC (shielded or not) in, or near, your home?

DC is more dangerous but produces so much less EMI. You write that you have some 300 m of underground cabling, which suggests that you require both in your home space. Why?

Some domestic equipment runs only on AC, while some runs an AC but then transform it and rectifie it to DC at various voltages.

Can you not  keep it simple and put all AC appliances, generators, invertors and etc at a safe distance from your home, with remote start switching. Build a Faraday cage (preferably in the ground) closer to the house and fill it with the transformers and etc to run everything else on DC.

There is a very lazy man living near me who worked out that he would expend a lot less personal effort if his washing machine was next to his rotary washing line. He put up a small shed and ran an extension lead and hose pipe. He now lifts the wet washing out of the washer straight onto the line. I can only admire his inventiveness, sadly I expect the time he saves himself is spent sat on his arse in front of the TV.

I know I have come late to this conversation and apologise as you have probably answered most, if not all, of these questions in previous posts. (I`m too lazy to read the last 1500 posts) :laugh:

Fiber is the future of all terrestrial communication and data control, it is very difficult to hack and immune to EMP from solar flares or nuclear blasts( the routing systems are not). It is however fragile, I suspect that this is why the military rely so heavily on satellite communications and why the US, the former Soviet Union and China are investing so heavily in space defense capabilities.

Keep well,
Bob




BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: DIY Low EMF 5 step Sine Inverter
« Reply #21 on: August 16, 2018, 05:12:09 PM »
Bob, I use 120VDC for my home, plus some 12VDC for controls, etc. Both filtered with -110dB from 10K to 2GHz military grade filters and in EMT conduit with compression fittings.  I don't use AC in the home because it's only useful function is for running motors, which I don't use in the home because of sound and the ELF magnetic fields.  The inverter is using my 120VDC battery bank/PV charge controller as the source, and it's output shares the same conduit as the shop and house feeds (as does my ST-3 output). 

I have 250 feet of conduit from the battery bank building to the shop/house, so 76 meters. The building was located on a hill which is ideal for a wind generator, but I found PV was more than adequate even in winter, here.

You mention DC as being less safe, when in fact, this is not true. For any given voltage, DC is about 4x safer, including the "can't let go" threshold.  This is well established.  I thought otherwise myself until I did some research.

There is no difference between AC and DC for power emission as far as higher frequency EMI. I have misled you - I am only talking about the EMI on the DC and inverter generated AC wires, now measured in my shop while the inverter is switched on.   The same switching power supply for computer equipment on my 120VDC system makes just as much EMI on the DC as it does on AC.  (I have a 7 stage LC filter for the computer gear.) The difference is that filtering of DC is much easier and more efficient, since capacitors don't heat/dissipate 60/50Hz power while trying to remove high frequencies. There is no limit on inductors, either.  You can add as much capacitance and inductance as you need, though the same EMC rules regarding stages for "woofer", "midrange", and "tweeters" apply.

I developed this low EMI inverter so that I could run via PV power on sunny days my well pump (1380W), which is far removed from the house and shop, and the washing machine(1100W), which is in the shop. I also wanted to test my design theory which I first applied successfully to a redesign of the electronics of the EL-SID circ pump (the only brushless DC circ pump at that time) about 17 years ago; that EMI is best controlled by slowing the slew rate of switching; that most of the losses of slow switching are compensated by reduced gate drive power.  That method is now being used by Texas Instruments in their latest variable speed motor drive IC's.  The industrial control situation has gotten so bad (EMI-wise) that finally designers are trying to address the problem at the source, instead of the liberal and expensive application of EMC band-aids everywhere else.  I've taken the same basic approach to a more extreme level; for very slow switching of power MOSFETs, you cannot just increase gate drive resistance. It requires some additional specialized circuitry. It also requires very careful selection of the MOSFET. Today, some MOSFETs are now rated for linear operation with full safe operating area data provided.  When I first applied the slow switching method to the EL-SID circulation pump, I had to test and use higher voltage MOSFETs since no linear data was available. Today it's much, much easier and some companies make great choices for slow switching that have very low on resistance (high efficiency and little heat).

It is unheard of for a 1500W (or even a 100W) inverter of any kind to produce dramatically lower EMI on the connected AC and DC cables than that produced by a typical 10 watt power supply. This one has less audible EMI at the attached cables than an LCD wrist watch. Those  offered commercially today will typically obliterate the entire AM band anywhere in or near the house since all the wiring radiates this EMI.  Early on in my development of my inverter, I decided to just disconnect the more extreme slew rate control circuitry to see what the emissions would be like.  The entire AM band was wiped out, within 8 feet of any connected wires.  So by design, I'm reducing emissions by over 90dB, with very little efficiency penalty.  I am sacrificing some THD, intentionally, but 8-9% THD is better than my ST-3 at 12% and it allowed me to take slow switching to an extreme. 

This is not your typical DIY inverter.





 

 
« Last Edit: August 16, 2018, 06:09:25 PM by BruceM »

BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: DIY Low EMF 5 step Sine Inverter
« Reply #22 on: August 18, 2018, 06:37:03 AM »
Ruined a metal cutting bandsaw blade yesterday cutting a gap in a 450W toroid core that I stripped of wires, etc.  Last time I did it by hand and it wrecked about 6 hacksaw blades.  The grain oriented silicon steel is very hard, and has a thin ceramic coating which eats blades.  I hoped my bandsaw would power through it easily, but NO, it was painfully slow.  I filled the gap with JB weld.

Today I wound 72 feet of 10 awg magnet wire on the core- 120 turns.  I've never used wire that big before and won't be doing it again.  No room in the center, doesn't look nearly as nice as the 12 awg one.  Next time I'll try multiple strands of smaller wire and start with a bigger core.  The 12 AWG one (same cut core, same number of turns) tested well at 14 milli-Henrys.  This should be close to that, but half the resistive loss.  Got it installed in the place of the 5mH Hammond choke with plenty of room to spare.

I got the Siemens relay delay unit and the power resistors this afternoon, and installed that in my power pedestal outside the shop.  So the 230- 120V step down transformer is now "soft start", with 40 ohms (5- 8 ohm 100W resistors in series) in series with the windings for 3 seconds, then a second relay closes and shorts the resistors.  Tested it briefly with the inverter and it seems to work fine.  Ran out of energy and sunlight so will do a more thorough checkout tomorrow.

I got some small toroidal 10uH inductors for the H-bridge filter boards yesterday. That's up next after some more testing.









G



ajaffa1

  • Hero Member
  • *****
  • Posts: 1725
    • View Profile
Re: DIY Low EMF 5 step Sine Inverter
« Reply #23 on: August 19, 2018, 01:14:17 AM »
Wow Bruce, looks like you are really pushing the boundaries on this. have you tried cutting the toroid with a four inch grinder? Should do the job but the heat generated might damage the ceramic coating.

Does cutting the toroid have a significant effect on its performance as a choke? I can certainly see that trying to wind that much wire onto an intact toroid would be impossible.

I am looking out for a cheap second hand grid tied inverter to play with. I intend to try rectifying the output from a 5 Kva generator to drive it. Solar by day, Lister at night.

Finding it very difficult to get any information on how much ripple an inverter will tolerate, I guess they are intended to run off PV or batteries so no one has tested them with rectified/smoothed AC.

Bob

BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: DIY Low EMF 5 step Sine Inverter
« Reply #24 on: August 19, 2018, 02:47:30 AM »
Gapped laminated toroidal cores are used for large inductance DC chokes.  Powdered iron toroids are used for values up to about 250 uF.  The gap is needed to avoid saturation, just as powdered iron is used instead of ferrites for DC power inductors.  Alas, my gapping is non-scientific, as the annealed toroid opens when it is cut so the gap varies widely between inside and outside. It has worked, though.  I will look into thin kerf 10" diameter metal cutting blades for my table saw as ruining a $26 bandsaw blade for a single cut is too expensive. In quantity, manufactured gapped torroid core chokes are very reasonable in price, but the onesy-twosy prices are beyond my budget, so I've been making my own from my collection of surplus toroidal transformers.

Alas, my testing today showed that even a jury rig adding a 3rd choke and two more 10K uF capacitors made no difference in the DC ripple measured at my shop on the 120VDC supply.  My new 10 awg wound choke is performing identically to the Hammond 5 millihenry choke it replaced for a range of loads.

Oddly, ripple is more than twice as high when the well pump is running instead of similar wattage of 230V heat lamps in the shop. I am at 15mv of ripple for the will pump. I must ponder this anomaly and test more to sort it out.  I may move my 3rd choke to the 0V line and see how that affects the DC ripple.

I did answer the mystery of the constant 20mv of AC ripple I found a few days ago; it was idiotic operator error.  There is a double pole switch for the DC near the outlet.  If the switch is off, the cable just acts as an antenna, and the DC and AC cables in the shop run parallel and into the same box on the wall.  I did it today, recognized the more jagged waveform instead of the 120Hz sine the inverter and DC filter makes.  My IQ plummets as I start to have temporal lobe abnormalities when I'm near the oscilloscope, etc., so I struggle to keep on track and get my planned measurements done.  I'd fire myself if I only I could.

Bob, I think you're right about the GTI being designed for relatively clean but variable volts/amps input.  Only testing of a specific model will tell the tale.  I would consider using a higher resistance choke such as a toroidal transformer operating in saturation (no gap, just use the existing windings). I have used them for chokes like this, and despite being in saturation they do work quite well.  My generator charger used two 1000W transformer primaries as chokes, a 10K uF capacitor between. No capacitor needed on the battery side, they already are a capacitor.  This got the ripple level down to 20 mv or so at 10 amps/146V. The voltage drop may need to be added to, as that will keep the GTI's maximum power point sweep from cyclically overloading your generator.

Even 10% ripple may be fine for your application...you'll just have to experiment.










ajaffa1

  • Hero Member
  • *****
  • Posts: 1725
    • View Profile
Re: DIY Low EMF 5 step Sine Inverter
« Reply #25 on: August 19, 2018, 11:08:52 AM »
Hey Buddy, sorry to hear you are a bit out of sorts, hope it clears soon, how is your wife recovering from her hospital trip?

Regarding the GTI I am looking for information on how much ripple they will cope with. I have a cunning/stupid plan to use my Lister CS 6/1 to run the ST5 I now have and use it to feed my grid tied inverter at night.

The output will need to be rectified and then smoothed to a level that is acceptable to the GTI. I believe that I can probably export energy to the gird after sundown by doing this, running on waste oil. The most electrically expensive part of my day is the evening while cooking food and watching TV, so running my CS on waste oil costs next to nothing and exporting the excess cuts my energy bill and carbon footprint.

Bob


BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: DIY Low EMF 5 step Sine Inverter
« Reply #26 on: August 19, 2018, 06:37:55 PM »
Taking a break from the pneumatic string trimmer while the compressor catches up.  Rain is rare and nice here, but damn, it makes grass and weeds grow with abandon.  I made a Gast 1-AM air motor string trimmer that I can use. 

I think Glort's right on using an induction motor as grid backfeed.  It's the simplest and most efficient.  But if you already have the ST-5 and GTI, making a bit of DC is a good plan too.  Big chokes or inductors also go by other names, such as "line reactors" or "reactor" if you are looking for surplus parts as I always do.  Just look at the milli-henry (MH) and amps rating. You'd like 5MH or more, and amps at greater than or equal to your max ouput.  For the GTI, check out the maximum DC input voltage, and the suggested range of voltages.  You must keep your DC peak below that max, with some margin.  If you do that, it may work oddly but is not likely to do any harm to the GTI, regardless of how much ripple you have on your DC.



The GTI obviously things you have connected it to a PV panel array. Some will periodically decrease and then increase the current draw to see where the optimum power production is. Then settle in at that level for a minute or so.  When it does this maximum power point sweep, it's expecting to see a voltage drop and total power drop when it's pulling too much.  By putting in some resistance (which could be in part via inductors) you will accomodate it.
Another solution would be to NOT use an AVR, using the harmonic windings only, letting voltage and frequency drop with the load of the MPP sweep. It will take some experimenting to find the simplest solution that will keep the GTI happy enough, without grossly overloading the Lister. A governor spring which allows more rpm sag on load might be just the trick.

Sorry you're having a bad one today, Glort. 

I'm sure to be laid up tomorrow after my string trimmer extravaganza.







BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: DIY Low EMF 5 step Sine Inverter
« Reply #27 on: August 20, 2018, 02:06:39 AM »
You can buy a meter for measuring inductance.  Usually the same meter will do capacitance measurement as well, since much of the circuitry is similar.  Probably some very cheap ones from China these days.

If the newer GTI's all hunt fast for the max power point, that makes some sort of resistance inline necessary. It must see a drop in voltage immediately as it tries to draw more power.
It would be time consuming to work that out (to keep it from constantly torturing the generator) but I do think it's doable.

Pity there isn't an open source GTI that could be modified to do something more civilized.

mike90045

  • Mendocino Metro
  • Hero Member
  • *****
  • Posts: 1594
  • Mmmm BBQ
    • View Profile
    • Mikes Solar PV page
Re: DIY Low EMF 5 step Sine Inverter
« Reply #28 on: August 20, 2018, 07:02:44 AM »
Some of the MPPT controllers designed to work with wind turbines (Morningstar and Midnight Classic) have settings where you can limit the hunting, or use a different power curve to get a stable harvest.

ajaffa1

  • Hero Member
  • *****
  • Posts: 1725
    • View Profile
Re: DIY Low EMF 5 step Sine Inverter
« Reply #29 on: August 21, 2018, 01:06:22 AM »
Glort is absolutely right about my skill set. My experience with advanced electronics is very limited. I`m good with cabling and motor control, I`ve built hundreds of computers but that just involves plugging bits into a motherboard.

Designing and building a custom GTI is way beyond me. Judging by how unreliable some of them are it`s also beyond some of the manufacturers.

I could just run my home off the generators at night but switching from grid to generator power resets all electronics in the house. If you have ever tried resetting digital clocks and etc you will understand the problem.

I could partially rewire the house so that water heating, cooking and other high energy appliances run off my SOM but I worry that my wife would find it too complicated.

Bob