An energy harvesting project

[JulesP]

Gary,

Having now acquired the video and watched it, I must say it’s a brilliant exposition on the really important aspects of these devices and also of the larger social context of where these devices sit.

Most valuable!

J

[JulesP]

Hi Rodolphe,

I have again summarized a range of issues under sections to make it easier for you and others to follow. I will gather any queries at the end.

My SSG Observations:

I have completed a 24hr run and two 12 hr runs of my SG device (strictly an SSG) to see if ‘radiant effects’ show. The 24h run was using the PCB version of the circuit and the other two were with a hardwired version (components connected with wire and connecting blocks) rather than PCB tracks. Here I was checking to see if the PCB format was somehow interfering with the radiant energy flow.

The graph below shows that again my voltages are not going above 12.5 for a 40Ah battery discharged down to about 5-10%Ah, so with very little of its charge left. It took most of the 24h to reach 12.5V and the supply current was around 0.25A. I don’t have a way to increase that and perhaps it is too low to charge well with spikes of only 130V.



So far then, the SSG has not demonstrated radiant charging in a clear way despite a few others seeing such with their builds. However, I have made progress on the charging curve differences, in fact, I think it is now 'half' solved.


Charging Curves:

The issue of why you are getting the curves you are, and up to 15V, has been perplexing, but I think I have part of the answer, or at least the charging profile is now very similar to yours, just that the voltage is still 'low'.

In going through the various possible causes of the problem, I looked at the connection from the HV output to the battery. As you may recall, I don't have an unbroken connection but go first to a connection stud where other related connections meet up (see pic), before going on to the battery. I wondered if the stud was somehow radiating away energy. This is a possibility given that this type of energy travels along the surface of the conductor and, since the junction is not covered in an insulating layer and presents a significant junk of metal, it may cause a 'disruption' to the flow. So I have done some measurements using a direct connection (see pic) with both the v4 and the BD1 (SSG) to see how it changes things.

I was pleasantly surprised that with the v4 (parameters shown on the graph) the curve shape was much closer to yours and steeper at the start. It didn't reach above 12.75 during charging but I think this is a major breakthrough. So I have now done an SSG test using the direction connection as in the pic below that. The first obvious change is that the rotor is now spinning at close to 700rpm compared to 350-400rpm before. The derived CoP is now 1.16 (60 mins left after charging and discharging). A distinct improvement













Next build:


The logic of going back to build the SSG has now proved fruitful in that the next battery of tests will refine what parameters work best and I will report on these in a few weeks. The obvious development is for me to now use five coils, as in my usual arrangement, but each with its own transistor. While I could make a 6-winding single coil (5 power and 1 trigger), as in the SG books, I will try 5 separate coils, and where one of them will have the trigger coil (probably use the SSG coil I have recently made for those two windings). While pulse synchronization might be an issue, the accurate machining of my rotor may prevent that. Besides I could easily instead consider the 4 additional coils as 'recovery' coils and the single and trigger winding as the main drive coil.

The idea that the five coils I have used up til now, being in parallel, might be dissipating their energy into each other before it reaches the output diode, has been suggested, and the next build will test that idea. Others have said that parallel coils work very well. With everything else the same, it will be a useful comparison anyway.

Here is the draft PCB then for v5 and where I am still waiting on any thoughts about how to drive the circuit with the PWM any differently than I have done up till now. The exposed copper output tracks give me the option to thicken them with solder for lower impedance.

Since the limit of images attached is five, I have attached the next 4 images to the following supplemental post.

Pic in next post


Solid state triggering:

The results from driving the SSG using a PWM are significantly worse than when using the prescribed trigger coil, so if I am going to include an SS option on the v5 board then I would need to ensure that it was being triggered in an optimum way. So one of the features of the next build will be the option to drive it with a PWM, as is done with the v4 replication build.

The general view is that triggering via a dedicated trigger coil works best and I have certainly got better results using that compared to a PWM trigger with the SSG PCB. In the few CoP tests I have done with it I got 0.7 with the trigger coil and 0.4 with the PWM. For some reason, the supply current with the trigger coil is substantially higher which accounts for the lower CoP. These CoP values will be better though, now that I am using a direct connection.

One very positive observation using the SSG is that any elevated voltages obtained after testing, do not evaporate as soon as I put a load on them with a discharge. They are not like the ‘phantom’ voltages (from high levels of surface charge) that result from using 2 - 4.5kV spikes from earlier tests. My growing understanding of battery conditioning is that it is possible to charge with a mix of 'hot' and 'cold' electricity to customize the battery's conditioning response. These details are emerging from the company with whom I am exchanging ideas etc. I'm sure in time some of these innovations will bleed into my updated manual and device for others to use.

Overall I don't think my SSG is best equipped yet to show radiant effects as much as an SG, but it is showing some traits that are better than the v4 (my one that is). In particular, the voltages rises are less 'phantom' and more load-sustaining, even if the resulting CoP is below or just above 1 at the moment. The 1.12 value I have so far achieved is more 'solid' than those based on high levels of surface charge. Things can only get better from here, and perhaps the peak voltage achieved during charging is not as important for CoP results as was previously thought.


Internal resistance:

The question of the battery's internal resistance is important, and indeed so is the impedance at critical points in the system. Having just acquired a new deed cycle 100Ah battery, I thought I would put my milliohm meter across the terminals (see pic) to see if I could do a direct measurement of the IR. I had hoped it would give me a reading directly but that doesn't seem to happen. I think the chemistry has to be active in discharge for the electrolyte ions to contribute to the internal resistance pathway.

Yes, one can use the CBAs system and I have the enhanced package for that, but the values are way too high for a fluid-filled battery and so presumably include all the cables and other resistances. I was told one can insert a value into the software to offset that but I haven't found it yet.


Pic in next post



Battery size:

With all these issues in getting the higher voltages, certainly, the battery itself may be another major factor. To test this I am going to pulse charge the brand new 100Ah one which has a substantial mass. Mass is turning out to be one of those key factors in OU observations. In effect, the bigger the battery the better the response.

With this in mind then I recently was introduced to the work of Ossie Callanan from Australia. His self-running system uses an 'accumulator' that is, in effect, a third battery. He argues that the single charging battery is not able to process all the pulses but that the accumulator bank of batteries (dead and heavily sulfated ones he uses) can absorb the pulses and then apply its response back to the single charging battery. Every so often the supply and charging batteries can be swapped over or the charging battery replaced with another and the energy from the swapped-out one used.

The logic behind this approach seems similar to the 'three' battery system JB and others have talked about. I expect that the mechanisms behind them are similar.

Pic in next post

Here is a graphic depicting the comparison with the standard Generator mode. Are these functionally the same, despite the obvious polarity swap and the coil being switched at the top instead of the bottom (with corresponding diode reversal)? If they are then it would be relatively simple to add an accumulator to the current setup to observe a 'three battery' system.


Pic in next post


Queries:

1. What is a typical supply current for one of your 15V peak charging cycles with the AGM battery?


2. Do you have any charging graphs for other batteries besides the X1 AGM battery?


3. From your observations, how many cycles do you think it takes to reach a satisfactory 'conditioned' status?


4. Any thoughts on the optimum solid-state coil triggering?


5. Have you found an adjustment in the CBA-enhanced software to allow for cable resistance with IR measurements?


I will report on parameter comparisons with the SSG in the near future.

Julian

[JulesP]

Here are the remaining relevant pics re the above post:

[pearldragon]

Hi Julian,

My SSG Observations:
It is not completely clear to me when you refer to what in your setup, I restarted the beginning of this thread now a couple of times, but still unclear which part of your results refer to which setup/graph, but I’ll try to answer as best as I can/understand:

Looking at the 2 charging graphs I see:
Setup 1
230811: 5 coils in parallel, PWM, Commonground/generator mode. I assume then no magnets. So I
assume no rotor is used here, just solid state.

Setup 2
230812: 1 coil, trigger coil (=bifilar wound?), commonground/Generator mode (so NO classic radiant charging???). How did you determine the +/- correct amount of windings for the trigger coil & main coil? I assume then a rotor is used with 5 magnets (have the other coils been removed to prevent interference?) I assume the ‘charging data using monopole charger’ information belongs to this setup.

-You already made a good start with simplifying things to understand parameters better by going back to the simple PCB you now use. In case you’re using 5 coils go back to one coil for now, but it seems you did that in setup 2.
-Discharging a battery to 5-10% of its full capacity is very deep discharging, I assume you’re using a battery that is made for this kind of deep discharging? (Although I have to admit I did more or less the same in my initial tests with my Bedini sg in classic/radiant mode).
-A 40Ah battery sounds like a lot for a single transistor… then also recharging takes long (single coil, bifilar wound). This is a potentially on of the reason why you’re not able to push to higher voltages.
In the 3^rd Bedini SG manual, page 26 under the heading “Classic SG baseline input measurements”, there is some mentioning about amp draw. (I do see in your setup 1 & 2 a 7Ah battery mentioned, that seems more appropriate.)
When you’re using PWM, you can play with the duty cycle to increase the power, and see if you can push the voltage higher. If you use a trigger coil, you can reduce the base resistance, or better make it variable (see the second Bedini SG handbook) -> How does your signal from the magnetic field collapse/HV pulse look like? When you use 1 coils with trigger winding in classic radiant mode -> When the rotor is at full speed, is it single triggering, or do you see multiple ‘shots’ magnet pass? If the latter is the case your base resistance should be changed so it does single shots per magnet pass. As can be read in the 2^nd Bedini handbook -> with a variable resister you can ‘speed up’ your rotor after changing this variable-resister value when the machine runs. (but it seems you’re not running in classic SG mode, so the last parts is probably not applicable right now).

Charging Curvers:
230811 starts to look more like mine. Why did you not let it run longer? To see if the voltage would increase?

Wow! Huge difference changing your connection, no?? Would LOVE to see something happening with ANY of my machines, but so far changing connections didn’t change a thing so far…

Next build:
I’d not go to using 5 coils until the charging profile issue has been solved in this mode, but up to you. If you want to go to 5 coils as you describe it with 1 trigger coil, it comes down to very accurate machining AND accurate manufacturing process of the magnets. This is not based on my own experience, but theoretical thinking for the same reason you mentioned: pulse synchronization.
If you want to increase output power, I’d go for the setup from the handbooks -> 1 coil, multi wound, + trigger winding.
I’m confused about the solid state triggering: So in your setup nr 1 you use PWM to trigger the transistor (as in the image of post 108); no rotor. In the V4 board you were using a Mosfet with gate driver. So you’re question related “to how to drive the circuit with the PWM any differently than I have done up till now.” Is that you’re wondering how to drive the transistor in a different way?

Solid state triggering:
“So one of the features of the next build will be the option to drive it with a PWM, as is done with the v4 replication build.” So the V5 will have an option for a Mosfet and Transistor.

“The general view is that triggering via a dedicated trigger coil works best and I have certainly got better results using that compared to a PWM trigger with the SSG PCB. In the few CoP tests I have done with it I got 0.7 with the trigger coil and 0.4 with the PWM.”
You mean with this new tiny PCB? -> With the V4 PCB you got COPs (single cycle) that were way higher.

“For some reason, the supply current with the trigger coil is substantially higher which accounts for the lower CoP”
This seems contrary to what you stated above. I would expect the sentence to be:
“For some reason, the supply current with the PWM trigger is substantially higher which accounts for the lower CoP”

Answers to questions
Question 1
It is unclear to me whether you refer to the my built of your SolidState (SS) setup, or my Bedini SG build and in which mode. So I’ll try to answer them all .
For extensive data and graphs, see attachment.
Summary:

Bedini SG Classic Radiant mode, Output battery: AGM A
Input: PSU input 12.3V @ 0.805A

Bedini SG Generator / Common ground mode, Output battery AGM A
Input S2 + S3 in parallel 13.22-12.87V @ 1.78-1.53A

Bedini SG Generator / Common ground mode, Output battery S1 (FLA)
Input S2 + S3 in parallel 13.29-12.93V @ 1.94-1.5A

Solid State Charger, V4 PCB Generator / Common ground mode, Output: battery LA1 (FLA)
Input: PSU 12V @ 1.57-1.36A

Question 2
I think this is answered in question 1

Question 3
This is the procedure I would do with a new battery OR when a battery has been sitting on the shelf for a while:
Do charge cycles with a REGULAR charger, while monitoring the charge profile. Discharge with CBA, somewhere between 8-20% out of the battery (but always the same chosen value). Repeat this process until your charging profile becomes stable. That means +/-same charging times & same max voltage where the charger stops charging. Depending on the battery this can easily take up 10-15 cycles… (NOTE: I needed to do it this way since I disconnected my regular charger after 10h of charging during the initial cycles when the battery was still not charged. Would I have left the charger on for longer, I might have needed way less cycles…)

Then start using an pulse charger (Bedini SG or solid state) and repeat; that means don’t change any variable until you get more or less stable results. I expect within 5 or 6 cycles.

Question 4
You’re the Solids state Guru here, not me . But as can be read in my other comments, not quite sure which situation we’re talking about.

Question 5
No, but also not looked for it. If you doubt the values of the CBA, I’d say contact WestMountain Radio (and keep me in the cc please ): They’re pretty responsive to inquiries.
Post 109 - Attachment.pdf

[pearldragon]

As a next step for me I was considering using a new PCB v4 build, bypassing anything I do not need for the basic functionality (e.g. the whole swapper circuit). But now I'm considering to start testing with that small PCB you're using at the moment as well...: Much simpler, and therefore also easier to compare results with your, and once we're a bit aligned -> build up to more complex stuff again.
What do you think about this?
May I trouble you for the manufacturing files of that PCB and a list (with links) of the components that you used?

(Still in the switching housing transition, will be very busy in the upcoming 2 months, but at least I can try to order the components already).

Best regards,
Rodolphe

[JulesP]

Hi Rodolphe,

I will respond to your longer post over the next week but to your last post then yes, I would be happy to send you the Gerber files for printing and the components list. However, I have one spare board left I think, in which case if you email me your address I will post it to you. But again, I have made a few revisions to the first board, like including an SG/GEN switch option and if you prefer I will send you the updated circuit and Gerber files for the latest version.

The components list will be emailed (I will need to update it with some useful links)

J

Good luck with the house upheavals!

[JulesP]

Hi Rodolphe,

Here are my responses to your responses to my earlier responses etc . . . . . .

It is not completely clear to me when you refer to what in your setup, I restarted the beginning of this thread now a couple of times, but still unclear which part of your results refer to which setup/graph, but I’ll try to answer as best as I can/understand:

Looking at the 2 charging graphs I see:
Setup 1
230811: 5 coils in parallel, PWM, Commonground/generator mode. I assume then no magnets. So I
assume no rotor is used here, just solid state.

Yes, just using the SSG (BD-1) circuit in place of v4 with PWM to see how it behaved. I hadn't wound the bifilar coil at that point.

Setup 2
230812: 1 coil, trigger coil (=bifilar wound?), common ground/Generator mode (so NO classic radiant charging???). How did you determine the +/- correct amount of windings for the trigger coil & main coil? I assume then a rotor is used with 5 magnets (have the other coils been removed to prevent interference?) I assume the ‘ charging data using monopole charger’ information belongs to this setup.

Yes, no SG mode as that is not working on my setup (due to the way I have connected the out - to the supply +, but should work on the revised board I have sent you the files for as I’m trying a different connection in the revised design.

-You already made a good start with simplifying things to understand parameters better by going back to the simple PCB you now use. In case you’re using 5 coils go back to one coil for now, but it seems you did that in setup 2.

The other coils are all disconnected but sitting there are lumps of conductor so will certainly have an effect on the rotor. One piece of very important information that I have just confirmed is that the extra coils in my setup, while not electrically connected to the BD1, are indeed making a substantial and positive contribution to the CoP. Since radiant energy gets everywhere in the system, the extra mass that the coils provide acts like the extra plate mass in a larger battery. A recent test with all the extra coils removed, and so with just the bifilar coil in place, came in at 0.5 compared to 1.05 with the extra coils (with no cores) in place. I will see if putting the ferrite cores back in increases this still further.

This implies that with my SSG, if it is running with just a single bifilar coil, will not achieve OU despite all the refinements of the parameters. However, I have not yet installed the larger rectangular magnets onto the rotor circumference to more closely mimic the original monopole motor design, even if my rotor is not a larger diameter bicycle wheel.

The upshot of these findings is that radiant energy can and does use any localized conductors and magnetic material to distribute its effects and so a suitable array of magnetic/conducting mass will enhance the overall performance.

This is similar to my observation with the v4 in SS mode, that when I pulled out the rotor the performance went down since the rotor magnets were enhancing the fields in all the coils being triggered by the PWM.

This is one of the reasons why I plan to use five coils instead of one with 5+1 windings, the extra ‘magnetic mass’ will contribute to the effect. I will still be able to use a multiple-winding coil with the v5. For example, I could measure the relative performance of one 3+1 coil, taking up three switched channels, and then the remaining 4 coil slots will provide 2 x 2 coils in parallel taking up the last two channels. This will give a mix of three fully synchronized pulses together with 2 ‘machine tolerance’ synchronized pulses; if you get my drift. This can be compared to 5 ‘machine tolerance’ synchronized pulses.

-Discharging a battery to 5-10% of its full capacity is very deep discharging, I assume you are using a battery that is made for this kind of deep discharging? (Although I have to admit I did more or less the same in my initial tests with my Bedini sg in classic/radiant mode).
-A 40Ah battery sounds like a lot for a single transistor… then also recharging takes a long time (single coil, bifilar wound). This is potentially one of the reasons why you are not able to push to higher voltages. [/FONT]

I took the 40Ah down very far for a particular type of test but it has recovered it seems and is back up to around 12.5V and ready for later runs when I compare the 7Ah with 17,18,40 and 100Ah batteries. I now have a deep cycle 100Ah battery that should prove interesting.

The larger batteries are said to respond better to radiant energy as they have a larger mass and mass appears to be important. As you say 0.25A is small and so will take days to charge up such a battery but I have no reasonable way to change the current without reducing the effectiveness of the BD-1 via the tuning. No doubt using additional coils in parallel with this circuit will increase the current and charging effect. Also the v5, with its 5 separate channels will likely draw much more current.

I’m using 7Ah at the moment to be able to conduct 2-4 tests in a day while I explore the effect of changing the many parameters. Once I have tied down the responses then I will increase Ah.

In the 3^rd Bedini SG manual, page 26 under the heading “ Classic SG baseline input measurements”, there is some mention about amp draw. (I do see in your setup 1 & 2 a 7Ah battery mentioned, that seems more appropriate.)
When you’re using PWM, you can play with the duty cycle to increase the power, and see if you can push the voltage higher.

Yes, I have played with the duty while using the PWM to increase current but again, the CoP using the PWM is substantially lower than using the trigger coil.

If you use a trigger coil, you can reduce the base resistance, or better make it variable (see the second Bedini SG handbook) -> How does your signal from the magnetic field collapse/HV pulse look like? If the latter is the case your base resistance should be changed so it does single shots per magnet pass. As can be read in the 2^nd Bedini handbook -> with a variable resister you can ‘speed up’ your rotor after changing this variable-resister value when the machine runs. (but it seems you’re not running in classic SG mode, so the last parts is probably not applicable right now).

I have adjusted the trigger circuit as proposed in the SG-3 book, adjusting it so that the rotor runs fastest and with the lowest current. As the trigger circuit is acting like the TDC adjustment in an old type engine, then we want the lowest current which here is around the 0.2-0.28A mark.

I don’t see any way to increase the current without reducing efficiency and I don’t see the need to. It will naturally increase when the coils require it. The whole setup seems to display a feedback loop from the battery back to the coils in that the state of the battery, where it is on its charging profile, and various other factors, all affect the current demand of the circuit.

When you use 1 coils with trigger winding in classic radiant mode -> When the rotor is at full speed, is it single triggering, or do you see multiple ‘shots’ magnet pass?

I have only done one divider/scope for each of the transistors. Both have the clear ‘h’ shape and maybe that is a good thing as it indicates some current/charge component instead of just a potential-only pulse?

Pulse comparisons.jpg

I haven’t had the circuit run in SG mode yet, only Gen mode but it would be interesting to see how the pulse train looked. I have adjusted the trigger pot for maximum rpm and minimum current as specified in the SG-2 book.

Wouldn't multiple output pulses with each magnet pass be useful?

Charging Curvers: 230811 starts to look more like mine. Why did you not let it run longer? To see if the voltage would increase?

This charging curve didn’t seem to want to go much higher but only a much longer test would show if it would. The voltage seems to plateau out at around the 13V level and increased charging just seems to lock in the charged status more - but other tests will clarify that.

How long were you charging your 12Ah AGM battery to reach 15V and at what supply current? Also, have you tried measuring the current in the + charging cable? When I used a clamp meter on it I got a negative reading! Is that ‘negative energy’? Lol

Wow! Huge difference changing your connection, no?? Would LOVE to see something happening with ANY of my machines, but so far changing connections didn’t change a thing so far&#8230 [/FONT]

Yes, and you are already using a ‘clean’ direct connection with the v4. I have been using the connection hubs/studs that seem to have disrupted the flow. I am soon to test a direct connection straight off the Drain and then without the cable insulation. This might show if insulated or non-insulated cable makes a difference and this will influence the v5 board design.

Next build:
I&#8217I'd not go to using 5 coils until the charging profile issue has been solved in this mode, but up to you. If you want to go to 5 coils as you describe it with 1 trigger coil, it comes down to very accurate machining AND accurate manufacturing process of the magnets. This is not based on my own experience, but theoretical thinking for the same reason you mentioned: pulse synchronization. If you want to increase output power and go for the setup from the handbooks -> 1 coil, multi wound, + trigger winding.

In the v5 plan, as I say above, I do have the option of say a 4 winding coil (3 power and 1 trigger) and using the other 2 channels with the remaining 4 coil positions (so 2 coils in parallel for each remaining channel). I will also be adjusting the magnetics by attaching larger rectangular ones to my rotor (with the SSG) to see if the flux linkage will be improved over my 20mm dia rotor magnets. Should be interesting.

I’m confused about the solid state triggering: So in your setup nr 1 you use PWM to trigger the transistor (as in the image of post 108); no rotor. In the V4 board you were using a Mosfet with gate driver. So you’re question related “to how to drive the circuit with the PWM any differently than I have done up till now.” Is that you’re wondering how to drive the transistor in a different way?

When I first assembled the BD-1 it was of interest to see how it responded to a PWM input as used in the v4 since I had not yet wound the bifilar coil. I also wanted to see how the 2N3055/MJL21194 responded to a square wave input with just one of my regular coils (the other were not connected to this board).

My question is that JB designed a solid state setup but did not share it widely so maybe there is a way to drive the SSG/SG type circuit that is better than the way I have done so far (PWM/driver chip and FET). If there is then I would like to integrate it into my v5 so it can be used in comparison to the trigger coil method.

Yes, I got bigger CoPs with the v4 but they have a more ‘phantom’ voltage quality where the energy would not be there on discharge. With the BD-1 the CoP is closer to 1 but the energy available upon discharge is more ‘real’. If I had measured Ah in and out I would likely have got lower CoPs.

This is the main reason why I decided to go ‘retro’ to the SSG to see how the charging was different. It certainly seems to be although I don’t know exactly how or why.

The primary issue of the CoP values with trigger coil vs PWM is the effectiveness of the resulting pulses on the battery. For reasons that are not yet clear using the PWM is not as effective as the trigger coil. For any one trigger method, a higher supply current will reduce the CoP but here there is something different in the way the trigger coil works with the transistor compared to how the transistor responds to a PWM trigger. It may be due to a ‘resonance’ phenomenon in the transistor as I have heard it said that these two transistors are the only ones with the right qualities for producing effective radiant charging. I have no idea yet what those qualities are.

It is unclear to me whether you refer to the my built of your SolidState (SS) setup, or my Bedini SG build and in which mode. So I’ll try to answer them all.

I was referring to the v4 for supply currents but it’s helpful to have others too. With your SG I presume this is a 7-power and 1 trigger coil winding as per the book? I expect the current values are typical and they are much larger than with my SSG with its 1+1 windings.

With the v4 these are in a similar range to mine although the duty cycle will make a wide range of options there.

The current demand can have wide-ranging consequences. Firstly you may achieve better and faster charging that will probably contribute to you getting the curve shape you have however, the higher current will drop the CoP (larger denominator in the energy division) unless you are electing a proportionally larger radiant effect.

Once again it’s finding the sweet spot with each variable. Coil voltage (load value) is particularly important here and where there is a higher chance of radiant effects anyway (SSG/SG) then small adjustments can nudge you OU. Hence the supply options I built into the v4 and which will also be in the v5.

Lowering or raising the load coil V to around 12V, down or up from whatever the supply battery happens to give you, can be significant (as can various other factors!) A cheap Buck converter can be worth getting (e.g.
https://www.ebay.co.uk/itm/314027833106) and similarly with a Boost converter. More to the point they help to maintain the voltage when the battery voltage is dropping over time.

This is the procedure I would do with a new battery OR when a battery has been sitting on the shelf for a while:
Do charge cycles with a REGULAR charger, while monitoring the charge profile. Discharge with CBA, somewhere between 8-20% out of the battery (but always the same chosen value). Repeat this process until your charging profile becomes stable. That means +/-same charging times & same max voltage where the charger stops charging. Depending on the battery this can easily take up 10-15 cycles… (NOTE: I needed to do it this way since I disconnected my regular charger after 10h of charging during the initial cycles when the battery was still not charged. Would I have left the charger on for longer, I might have needed way less cycles…) [/FONT]

Thanks for sharing your conditioning process. Why do you think that you shouldn't start IPC (inductive pulse charging) straight away with a new battery? Is there an essential need for regular charging first?

[Continued in next post due to character limit]

[JulesP]

[continued]

CoP methodology

I have attached the CoP methodology I have been using with my SSG bifilar setup. Using this I am testing to see the effect of a wide range of variables including the transistor, diodes, start voltage, charging time, rest interval, charging + connection, use of insulated or uninsulated wire, rotor magnets, number of coils and core type - and the type of coffee I drink!

The findings will inform the v5 design currently in development, which is basically a form of SG, and which will most likely be the basis of an ‘official’ research study starting early next year for which I am now preparing.

The upshot of all this is that rotor switching with a trigger coil has significant performance advantages over SS at the moment since the optimum way to trigger the transistor with a square wave is not yet fully clear. I understand that JB was disclosing more and more on this in his later years, but this process was muddled by ‘personal’ issues in connection with those who took information on this topic without his permission.

In other words, the solid-state story, with regard to my work, has yet to be fully written and explored. Meanwhile, the rotor-based story is unfolding in a very positive way and will provide valuable material and understanding regarding how radiant energy behaves and can be directed toward a self-running system.

Bear this in mind if you choose to use the BD1 board since it will give the best results with a rotor setup, such as with your own SG, and not with the v4 stand-alone coil setup.

I’m also sure that my liaison with the company that works in this area will also positively contribute to the solid state story in due course.

It seems our posts are getting longer and longer!

Regards,

Julian

[pearldragon]

Hi Julian, in response to your post #112:

“The other coils are all disconnected but sitting there are lumps of conductor so will certainly have an effect on the rotor. One piece of very important information that I have just confirmed is that the extra coils in my setup, while not electrically connected to the BD1, are indeed making a substantial and positive contribution to the CoP. Since radiant energy gets everywhere in the system, the extra mass that the coils provide acts like the extra plate mass in a larger battery. A recent test with all the extra coils removed, and so with just the bifilar coil in place, came in at 0.5 compared to 1.05 with the extra coils (with no cores) in place. I will see if putting the ferrite cores back in increases this still further.”
That is huge… did you repeat both test, to test for repeatability??

This implies that with my SSG, if it is running with just a single bifilar coil, will not achieve OU despite all the refinements of the parameters. However, I have not yet installed the larger rectangular magnets onto the rotor circumference to more closely mimic the original monopole motor design, even if my rotor is not a larger diameter bicycle wheel.
I’m not sure if that conclusion can be drawn so quickly… That term might apply for your setup, but not necessarily for somebody else who’s setup is slightly different (different magnets, amount of magnets, battery etc) . (With SSG you mean common ground/generator mode, I state it here for other readers; the terms start to pile up referring to the same circuit . I prefer to stick to the terms used in the handbook to avoid confusion.)

This is similar to my observation with the v4 in SS mode, that when I pulled out the rotor the performance went down since the rotor magnets were enhancing the fields in all the coils being triggered by the PWM.
Ok, this is very interesting information too. Are you saying that in all the reports you published, regarding the V4 board, you had the rotor in the middle (doing nothing) and the coils around it? If so, than I should strongly consider changing my coils setup to something similar with magnets (rotor) in the middle and redo my tests…
You loose me completely again in the next part haha. Let’s do a sentence by sentence interpretation and see if I got it right:

This is one of the reasons why I plan to use five coils instead of one with 5+1 windings, the extra ‘magnetic mass’ will contribute to the effect.
-You mean here 5 coils, bi-filar wound, instead of 1 coil with 5 power windings and 1 trigger winding (=5+1).
The extra magnetic mass comes here from the cores?
I could measure the relative performance of one 3+1 coil, taking up three switched channels, and then the remaining 4 coil slots will provide 2 x 2 coils in parallel taking up the last two channels.
-1 Coil with 3 power windings (using 3 of the active devices) and 1 trigger winding , triggering the 3 active devices.
-2 sets of 2 coils in parallel, triggered then by the trigger wire of the 3+1 coil?
This will give a mix of three fully synchronized pulses together with 2 ‘machine tolerance’ synchronized pulses
-Based on this sentence I would understand that the 2 sets of 2 coils in parallel are NOT triggered by the 3+1 coil, but botg sets have one of their coils bi-filar wound.

The larger batteries are said to respond better to radiant energy as they have a larger mass and mass appears to be important. As you say 0.25A is small and so will take days to charge up such a battery but I have no reasonable way to change the current without reducing the effectiveness of the BD-1 via the tuning. No doubt using additional coils in parallel with this circuit will increase the current and charging effect. Also the v5, with its 5 separate channels will likely draw much more current.
In case you’re interested, I think I still have a 8 channel PCB lying here from TeslaGenX. (but you might be just as quick making it yourself, and making it more versatile while doing so.)

I have only done one divider/scope for each of the transistors. Both have the clear ‘h’ shape and maybe that is a good thing as it indicates some current/charge component instead of just a potential-only pulse?
I need to think this through a bit better to confirm it, but the fact that you see the h-shape defined with the transistor and just a peak with the MOSFET, might also have to do with the speed with which the MOSFET operates. Maybe if you zoom further in you’d see something similar in the MOSFET pulse/signal, but again need to think this through a bit more. Gary can probably give an answer with a bit more reason behind it .
The image under this text does not work, can you repost?

When I first assembled the BD-1 it was of interest to see how it responded to a PWM input as used in the v4 since I had not yet wound the bifilar coil. I also wanted to see how the 2N3055/MJL21194 responded to a square wave input with just one of my regular coils (the other were not connected to this board).
My question is that JB designed a solid state setup but did not share it widely so maybe there is a way to drive the SSG/SG type circuit that is better than the way I have done so far (PWM/driver chip and FET). If there is then I would like to integrate it into my v5 so it can be used in comparison to the trigger coil method.
I completely understand your question now. During the week suddenly I understood what you meant and had the same question myself (if that was the correct way of doing it). I’ll put it on my list to discuss this with some of my friends who are into electronics, and will come back on this (I just don’t expect it to be very soon….)


Wouldn't multiple output pulses with each magnet pass be useful?
This is another question that Gary can answer probably better than me. What he told me is that in Classic/radiant mode one should aim for a single shot per magnet pass, which happens if the rotor gets close to its max speed and is tuned at least half decent. (So it can do multiple shots at starting up.).
I don’t recall sawing multiple shots ins generator/common ground mode…

This is the main reason why I decided to go ‘retro’ to the SSG to see how the charging was different. It certainly seems to be although I don’t know exactly how or why
Still no strange response from your PSU? With the small BD-1 boards?

I was referring to the v4 for supply currents but it’s helpful to have others too. With your SG I presume this is a 7-power and 1 trigger coil winding as per the book?
8+1 --> 1 power windings, 1 trigger winding

Coil voltage (load value) is particularly important here and where there is a higher chance of radiant effects anyway (SSG/SG) then small adjustments can nudge you OU
In the tests I did with your V4 board, so far I have NOT changed this value… kept it in all tests at 12V; I first wanted to see a change in COP by changing the frequency.

Why do you think that you shouldn't start IPC (inductive pulse charging) straight away with a new battery? Is there an essential need for regular charging first?
The thing is that you at the moment you do not know if any of your systems are able to generate enough power/input for the battery to be charged to higher voltage levels… I’f you’d know that for sure, there would be no reason why you could not do it with a pulse machine/charger. So with that uncertainty there, I’d try to eliminate that in the beginning -> first charge with regular charger till you see nothing changes anymore -> repeatable results. Then you know at least that you battery is as far conditioned as your regular charger can do.

Best regards,
Rodolphe

[pearldragon]

In response to your post #113:

Bear this in mind if you choose to use the BD1 board since it will give the best results with a rotor setup, such as with your own SG, and not with the v4 stand-alone coil setup.
Well, I’m still a couple of question down the path: why does my V4 board perform SO different than your one, that in any of the tests I did I did not go over 0.45 COP (Ah). But yes, comparing that specific battery with my Bedini SG commonground/generator mode, the latter would perform better --> 0.55-0.62 COP.
Going back to an easier system like the BD-1, might give better starting point - > hopefully make our results more similar.
Although having those magnets in the middle, and rearrange my coils like you have them ( in a circle around the magnets/rotor) is something that has my interest too...

It seems our posts are getting longer and longer!
This one proves the opposite

[Gary Hammond]

Hi Rodolphe and Julian,

Wouldn't multiple output pulses with each magnet pass be useful?
This is another question that Gary can answer probably better than me. What he told me is that in Classic/radiant mode one should aim for a single shot per magnet pass, which happens if the rotor gets close to its max speed and is tuned at least half decent. (So it can do multiple shots at starting up.).
I don’t recall sawing multiple shots ins generator/common ground mode…

With my SSG when running in radiant mode I get the best charging with TWO pulses per magnet pass. And in common ground (generator) mode I get the best charging of all, and it is always at one pulse per magnet pass. The wave forms are different between "radiant" mode and "common ground" mode. I only get single pulses in "common ground" mode even at start up.

Here's a link to one of my old posts where I discuss this with Alvaro. https://www.energyscienceforum.com/f...7843#post27843

Regards,
Gary Hammond

[pearldragon]

Hi Julian, Gary

Thanks Gary for your input and corrections.

In response to your post #113 Cop measurement method:

We talked about this topic a couple of times. I do appreciate your take/point of view towards this topic and understand why you do it the way you describe.

It is not way that I myself prefer for the following reasons:

1) In your method you keep the charging time constant. And wherever you end up as a final voltage level (after letting the battery rest) you extrapolate/interpolate to the voltage level you started with before discharging.
In my opinion you can get inaccurate COPs, since if your charging process stopped in the ‘steepest’ part of your of the full charging cycle, this will give an higher COP as can be shown on page 69 of 81 in your manual.

2) The battery might respond very different by discharging different portions out of it. E.g. I get a different COP if I charge 1Ah out of my battery, than when I charge 0.5Ah out of it…
For the above mentioned reasons my charge cycle time is always a result, and never a set value. My set value is the voltage to where I charged the previous charge cycle: Then I know I filled up the battery to exactly the same point after taking charge out of it during the discharge step.

Best regards,
Rodolphe

[Gary Hammond]

Hi Rodolphe,

1) In your method you keep the charging time constant. And wherever you end up as a final voltage level (after letting the battery rest) you extrapolate/interpolate to the voltage level you started with before discharging.
In my opinion you can get inaccurate COPs, since if your charging process stopped in the ‘steepest’ part of your of the full charging cycle, this will give an higher COP as can be shown on page 69 of 81 in your manual.

That's not what I did at all! I charged until the receiving battery reached 16.0 volts under charge and then stopped it and recorded the time it took. This turned out to be 35 minutes at 1.6 amps current draw from the primaries. This is plainly stated in post #44 which I link to here. https://www.energyscienceforum.com/f...9921#post19921 I then noted the resting voltage one hour after I stopped the charge and it read 12.79 volts which was still a little higher than it would be if it sat over 24 hours. That would normally be 12.69 on that particular battery at the time this was done (2-15-2014).

This was in generator (common ground) mode which will keep raising the voltage well past the normal fully charged point of 15.3 volts and produce off gassing. I've taken it clear up to 17 volts and it still drops back to the same voltage range after resting. Because of this it looks on the graph like it is still in the steep range, but if you look closely you can see that the charge curve is beginning to flatten out when I stopped it at 16 volts. I didn't want to take it any higher because it was off gassing instead of charging.

Regards,
Gary Hammond,

[JulesP]

Gary, in post #117 Rodolphe is describing my method of deriving CoP and not your one.

[Gary Hammond]

Gary, in post #117 Rodolphe is describing my method of deriving CoP and not your one.

Oops. Sorry. I misread the post. My bad!

Gary Hammond,