Need expert feedback on my SG build based on book 2 page 47 and book 3 page 23!

[psiegers]

Hi all,

With this thread I'd like to start obtaining some expert feedback on a Bedini SG build I built a couple of months ago - actually my first build.
Basically I need advice on how to increase output - as it charges at a very very slow rate.
Much more information will be revealed in a couple of subsequent posts, to give a very detailed context.
Thanks a lot in advance for any help you can provide!

Pieter

[psiegers]

First of all, the schema, of course, with all important details.

[psiegers]

Then some pictures of my build.
First 5, then 2 more follow.

[psiegers]

The other 2 pics.

[psiegers]

Ok, time for a couple of next posts.

I've built this SG all by myself.
I bought a second 26" wheel and glued 24 C8 magnets on it, and together with the shop owner made it as true running as possible.
As the wheel has 28 spokes, there actually are still 4 open spaces for more magnets - which may not have been a good decision but at the time when I thought - I do still have 4 flat very strong neodymium magnets I bought several years ago and wanted to mount them to see what would happen - but was a bit frightened afterwards as they could well damage the device in several ways and besides that fly around the room causing some real trouble.
I applied some wrapping tape over the magnets to prevent them from being thrown in the air - which already proved to be a very good safety measure!

I built the frame from pretty sturdy wood - mostly based on what the 1st SG book showed me.

Then I proceeded with the coil.
I wound it by hand - first I wound three twisted wires, one 23AWG trigger wire, and two 20AWG for input and output - all about 150 feet long.
After that I added 6 more 20AWG wires also 150 ft, to be able to have more 'juice' at the output, at least that is what I thought I'd get.

Then it was time to assemble the electronic components.
This is a completely self build on practice board with thick wires to connect the high current paths, and every part that required a cooling device got one.
After verifying connections time after time, it was the most daunting task to get it all done well.

When all was assembled and separately tested as much as possible with a power supply with automatic current limitation, I initially ran the machine - that was a very special moment!
When running the machine only the 100 ohms resistor got a bit warm - but it started to run with still some serious wheel friction.
So I decided to concentrate on the wheel first, and found that loosening the bearings helped quite a lot.
Now the machine would run by itself and accelerate slowly to around 100 RPM.
The input current was around 0.5 amps - not so bad I thought - but the wheel was still turning way to slow.
So, I decided to add one 100 ohms resistor in parallel to the already existing one, so the resulting resistor is now around 50 ohms.
This resulted in a quicker turning wheel (around 180 RPM) and with an input current of around 0.7 amps.

OK, so next posts will be about my first experiments with the two worn out Trojan batteries I want to rejuvenate.

[psiegers]

Ah, about the coil, it has an iron core which I had harvested from an old transformer.
I inserted several packages of laminated iron with a surface that covers the whole C8 magnets passing by.
Peter had this idea when he built the pickup coil described in the 3rd SG book.
After that, I wound the twisted copper wires- see the pictures below.
The current air gap is small but I haven't done any measurement based on distance.

[psiegers]

OK, so now the batteries.
They are two identical Trojan 27TMX Deep Cycle at a rated C20 capacity of 105 AH @ 20 hours.
This means that when this battery is fully loaded and in very good conditions (read new) it should be able to deliver 105/20 = 5.25 Amps for 20 hours.
And then the cell voltage should not have fallen under 10.5V.
Well, I think that my two batteries were pretty good when they were brand new - but I never actually tested C20.
Now, around 5 years later they are almost completely powerless - they charge but too quickly, and therefore they discharge just a bit of juice.
Nothing of the energy I put in is actually stored.
According to Peter, I understand this is because the plates are covered with too many layers of sulfate ions that combined with Pb.
The problem is that with conventional loaders and also with the SG in specific, I can't get them past 12.6 Volt.
The process of rejuvenating is to gradually remove the sulfured layers from the plates, bring the ions back into the electrolyte and while electrolysis takes place recover the original chemical state of the battery.
That is what I'm trying to achieve with the SG, but it doesn't work.
I'd like to receive some expert advice on what am I doing wrong, or what I can do to increase the charging voltage on the charged battery up until 15,3V or close by so that point P in the graphic is reached and the rejuvenation process can begin.
A picture of one of the batteries was already included above in post #4.

[psiegers]

OK, if anyone reads this thread and has (or knows somebody) a Bedini charger 10A12 that I could buy please let me know!

[Gary Hammond]

Hi Psiegers,

OK, if anyone reads this thread and has (or knows somebody) a Bedini charger 10A12 that I could buy please let me know!

I think Teslagenx still sells these. I don't know if they keep them in stock or only build to order. You can contact them at this link. http://www.teslagenx.com/chargers/tx...egory=chargers

Gary Hammond,

[psiegers]

Thanks very much for that link Gary!

[psiegers]

A little bit about the capacitor.
I bought this one: https://www.mouser.com/ProductDetail...tCbtIA4gpQCA==
It's a big Kemet 130000 uF 100VDC one, with an ESR value of 6mOhms.
My capacitor discharges around every second in a very short instance.
How? The schema from post #2 shows all the details.

With one of the worn out batteries connected, the voltage when not discharging rises up quickly to about 13,8 V, and when discharged, it quickly drops to around the current battery voltage.
I do see voltage wobbling at around 1V, and it does rise to about 12V but never passed it - even not after having charged this battery for three days.
Using a conventional charger the battery charges but way too quickly but voltages behavior looks acceptable - the battery does not seem to have severe damage.

I'm pretty sure that using this capacitor the system should be able to push all its harvested energy into the battery with minimal losses.

Note: today I tried a brand new car battery which is around 85 Ah at C20 rate.
I discharged it yesterday for about 6 hours with 2 car lamps at 4,3 Amps.
Final rest voltage this morning: 12,47V.

When charging with the SG connected and running full speed, I see the voltage is not climbing, it is actually dropping - at about 0,01V per 10 mins.
This seems strange to me - I've checked the capacitor charge, and discharge - it does a bit of discharging into the battery, although it is very little.
So I am starting to think that my output circuit has some flaw somewhere - the incoming energy is seems too small to charge the battery.

Anyways, I'll leave the SG running for the rest of the day and see what happens.
It is pretty clear I definitely need something better for measuring the output.
Any comment or tips to analyze this behavior are very welcome!

[Gary Hammond]

Hi Psiegers,

With one of the worn out batteries connected, the voltage when not discharging rises up quickly to about 13,8 V, and when discharged, it quickly drops to around the current battery voltage.

A cap discharge circuit should push the cap up to approximately double the battery voltage (24 to 28 volts) and then discharge through the battery stopping about 2 to 4 volts above the battery voltage. It should be able to do this 2 or 3 times per second. As the battery rises in voltage, the cap voltages should rise with it.

The 105 AH batteries you are using are too big to bring all the way up to 15.3 volts with the SSG built by the book. I have similar batteries and have trouble getting them over 14.6 volts. Even my 10A12 Bedini charger has trouble getting them over 14.9 volts.

You might be interested in reading all the posts in this thread for reference. https://www.energyscienceforum.com/forum/alternative-energy/john-bedini/bedini-sg-official-monopole-forum/bedini-monopole-1-advanced/2294-charging-3-105-ah-batteries-with-ssg

Gary Hammond,

P.S. ---- You might be interested in this thread as well, especially post #7. I was using 13AH (U1L) batteries in this thread. https://www.energyscienceforum.com/f...nced-ssg-build

[psiegers]

Hi all, next update.
Thanks Gary for your comments - I did read up on the threads you mentioned, and decided I needed to have a closer look at what my SG build does in practice.

I'll first attach a picture of the connection diagram I drew quickly, and some temporary values I saw on my V and A meters (all analog) and scope.

I've switched to using two brand new car batteries - they are smaller than the ones I try to rejuvenate - specs say 85Ah at C20.
I use one as the run battery, and the other is the charge battery.

The voltage on the cap is around 13,55V - 14,14V, with the battery voltage being 13,08V.
I have the 555 switch running at about 1 second interval, with around 40 msec pulse width according to the scope image.
Meaning the difference is very little - hence, very little energy is transferred from the cap to the charge battery, and also in very little time intervals.

The scope image drawn shows the scope image, with the cap voltage discharge/charge drawn against the 555 pulse on base of T2.
The voltage difference is only about 0,4V at maximum - I now understand this is way too small to get such a big battery to rise its voltage even over a long time, because of losses.

So then, let's sum up the components in the output circuit that have a possible negative impact on the SG capacity to charge the battery.

1.There is the 7812 voltage regulator. It constantly draws current from the incoming energy. This represents actually a design flaw - when the cap voltage rises above 34V this component will simply die. I can tell because it happened twice already - and it took the 555 and some other components into their grave as well.
2. I have an ampere meter constantly connected in the negative wire to the charge battery - this low impedance also reduces charging capacity.
3. I have a volt meter constantly monitoring the charge battery - these are supposed to have a very high impedance but it still affects the charge capacity because of the supplied voltage spikes to the capacitor which are then translated in very high current spikes but still with considerable voltage spikes - hence, a loss.
4. Fourth, there is the series resistance of the two MOSFET - that should be very low but still, there's a small loss.
5. I have a low ohm switch in between the capacitor and charge battery, to be able to see cap charging without charge battery connected - but only for very short intervals!
6. With respect to heat generation, none of the components in the output circuit become warmer than the surrounding ambient temperature.
7. I have used very thick wiring for the high current paths - very small losses there.

Some of these losses are present all time, some of them during the 40 msec pulse width.
I read that amp meters should not be connected. So I'll add switches to shortcut them during tests.
About the other losses, I'm not sure what to do about it - at least not yet.

OK, I mentioned I have this switch between the cap and the charge battery, so when in disconnected state, it takes around 15 seconds for the voltage to rise to around the recommended voltage of 24V.
From your comments Gary - that you'd expect this to be happening around 2-3 times per second, my measured time is actually a very long time, I acknowledge that, but the cause may well be the big capacitor - it takes time to load that thing with little input.
Anyways, it does generate pretty high current spikes on the amp meter when the switch is in connected state, with pulses up to 5A (analog amp meter):
But yeah, having a system that waits like 15 seconds for each discharge into the charge battery, this is not practical in any way, even though the voltage pulses reach the positive terminal at any moment.

So, all in all, it's clear that even with the simple measurements taken, I think it is useless to continue expecting I can charge big batteries with this version of the SG.
But, looking at my schema, what could be the best next step to get more juice on the output and be able to charge a 105Ah battery?
Or should I just consider buying a 10A12 charger?
I could do that of course - I'm considering acquiring one - but I'd like to continue this SG build and improve on it step by step and learning along the way.

Have a good one all and take care.

[Gary Hammond]

Hi Psiegers,

From your comments Gary - that you'd expect this to be happening around 2-3 times per second, my measured time is actually a very long time, I acknowledge that, but the cause may well be the big capacitor - it takes time to load that thing with little input.

Your 130,000 uf cap is a lot bigger than it needs to be. Surprised it doesn't blow the two FETs that dump it. My cap dump uses 66,000 uf (three 22,000uf in parallel) capacitance discharging thru 4 FETs.

I didn't have very good results from the cap discharge unit I built according to the book. After blowing it up twice, I built a completely different unit which is discussed here. https://www.energyscienceforum.com/f...age2#post39754

But, looking at my schema, what could be the best next step to get more juice on the output and be able to charge a 105Ah battery?
Or should I just consider buying a 10A12 charger?
I could do that of course - I'm considering acquiring one - but I'd like to continue this SG build and improve on it step by step and learning along the way.

Build a bigger (multi coil) SSG and a larger, more efficient cap dump for charging those large batteries. And the 10A12 charger is always a good investment.

Gary Hammond,

[psiegers]

Hi Gary, thanks for your helpful comments.
I assume you have your capacitors in parallel so that the values sum?

OK so I have read up upon some of the mentioned threads, and some or more of the latest ESTC conference, very interesting stuff and a lot going on these days - just wonderful.

And, I also browsed some EnergeticForum threads and found this one by John Koorn.
The mentioned thread doesn't have any detailed information nor answers but from the video we know a couple of things:
- it shows two 8 transistor boards used
- it shows an 81 23AWG strand coil (with air core)
- 1 strand for the trigger circuit (Bedini SS Oscillator with an off/on switch which drives the 16 transistors in turn, at around 952 Hz - 1.24 kHz at the end of the test)
- 4 strands per transistor, totaling 64 strands, and then 2x8 strands that go directly to the charge battery positive
- capacitor charge circuit shown but NOT being used (meaning the charge battery is being charged with the wrong type of electricity (inductive) - the cap charger should be used to charge the battery with capacitive electricity which is compatible with the output of general available battery chargers)
- input batteries - 2 of each 50 Ah, 12V
- charge battery - 12V not clear how much Ah but the voltage was rising very quickly!
- circuit was in common ground mode
- John comments that in his opinion this circuit can charge much bigger batteries (so again how much Ah was the charge battery?)
Qs:
- why an 81 strand coil was used? I would think using a thicker wire for each 4 strands of 23AWG would simplify this coil?
- Exact circuit of the Bedini SS Oscillator?
- Can 105 Ah next to dead batteries be rejuvenated with this circuit?? There's only one way to find out

I guess I'll be preparing for a whole new Bedini SG build and going the SS way from now on.

Have a good one all.