No announcement yet.

An energy harvesting project

  • Filter
  • Time
  • Show
Clear All
new posts

  • #91
    Hi Julian,

    Thanks for the update/info to solve the spike issue with the amp meters. Will implement those caps (at some point) in my device as well (currently blew up both my meters ).
    I want to respond a bit more elaborate on your post, also on some previous ones, but still looking for the time to so, so for now I just share some of my initial ‘quick and dirty’ results.

    COP Tests
    Quick and dirty, because I wanted some ball park COP (Ah in versus Ah out), numbers and as long as they are far below 1 and/or, not a noteworthy change compared to a previous measurement I don’t care too much about the accuracy yet. So to explain what I mean with that/how I performed my measurement so far:
    -Discharge 1Ah out of my output battery
    -Next day charge with the Solid State Charger (SSC)
    ---Noting every 5 tot 10min the amp reading on the display of my power supply (see image below). (this measurement is probably a bit inaccurate, still have to start using my recording multimeter)
    ---Record the rising voltage in my output battery with the CBA
    ---Stop charging at 14.30V (since that is where my conventional charger charged to as well
    -Wait for an hour (while recording with CBA)
    -Then start at step 1 again Discharge 1Ah.

    For the calculation of the Ah the PSU put in I took the amp reading @ +/-60% of the charge time, and multiplied that by the total time of the run; as mentioned this is very rough measurement.

    Generally I would do one charge + discharge a day, except for the first tests (ID 230528 CM & 230530 CM), there I did two full charge/discharge runs in 1 day.

    Test IDs 230528 CM / 230530 CM / 230532 CM / 230534 CM are done with IRF840.
    Test IDs 230601 CM / 230603 CM / 230605 CM are done with STP20N95K5.
    Rest of the settings can be found in the attachments.

    Conclusions/follow up actions:
    -For test ID 230528 I did not discharge 1Ah out of the output battery the previous day, but 4 days ago. The COP is therefore +/-10% higher then when I discharge the previous day. This I noticed as well with the Bedini SG, sometimes this difference could be even far more than 10%.
    -With currently selected settings no significant difference between IRF840 and STP20N95K5.
    -I was planning to continue testing with the STP12N120K5, then STW12N150K5, finally STW12N170K5, but reasoned that finding the right frequency might be more important to figure out first, so will probably focus on that as a next step.

    Peak HV pulse of active devices
    I re-did the measurements, this time a bit more in a reproducible manner so I could track the increase of peak voltage over a couple of minutes and comparing the different active devices. I included my measurements for the IRF840 / STP20N95K5 / STP12N120K5.
    UPDATE 2023-06-05 -> Replaced attachment of active device measurements, contained a fault: 3rd test was not done with STP20N95K5 but with STP12N120K5.

    Best regards,

    Click image for larger version  Name:	Post 91 - PSU.PNG Views:	0 Size:	27.4 KB ID:	74219
    Attached Files
    Last edited by pearldragon; 06-05-2023, 12:05 PM.


    • #92
      Hi Rodolphe,

      It's good to see that you are starting to make measurements.

      The 0.5 value you have got so far is fairly consistent with standard electrical behavior and efficiency for converting the supply to HV pulses. Although strictly speaking a CoP value is a ratio of energies (energy out / energy in) and not Charge (Ah), it will be interesting to see how your figures might change when if you use the method I described in my first interim report (CoP), in other words using stabilized battery voltage as the focus instead of Ah in and out. Given that I have yet to achieve any net gains under load producing a CoP>1, with HV pulses or Cap Dump discharges, perhaps using Ah in fact gives a more accurate and meaningful result? If so then that leaves the burning question of why the charged batteries show very elevated voltages after pulse charging and resting.

      Your HV peak measurements are generally to be expected although it is interesting that on a different day you got a couple of hundred volts higher with the 95K5 device. Once I had a reading for a device, I would take that as a reference value as the divider can drift with temperature. Was the ambient temperature very different on those two days? In effect, I would tweak the divider adjustment knob to give the reference reading when doing subsequent readings.

      I am in the middle of trying 'pulse combining' (see attached) which has been difficult to achieve in that, for some unknown feedback reason, my FET driver chip would be destroyed within a few minutes of pulse mixing. I have now bypassed it for the purpose of these experiments since unless I test this idea and method of working, I will not have a complete picture of what might improve results.

      Generally speaking though, so far I have not found the charging battery to rise above its starting voltage when swapping is involved and the elevated charged status is then used as the supply. It is possible then that at the moment, for reasons not yet clear 'radiant' effects are not manifesting to a noticeable enough degree, even though all my CoP tests, using battery voltages, would indicate otherwise. However, I have yet to try a 3.2kV and a 4.5kV IGBT device to see if almost tripling the spike HV will induce some as yet unseen effects. These are the voltages that Babcock uses to go beyond the 'ether tidal region' as he describes it. However, he has not responded to my communication with some queries based on his 2022 talk.

      So after these experiments, if I see no improved results, I am going to write up my load experiments and then follow the advice of a contact of mine and reconfigure my system as a more basic SG-type device to try and get some key radiant effects and then change one thing at a time to 'ramp it up in size and power over time. This will perhaps be the optimum way for me to find out whether, for example, impedance is the major blocking factor and where it matters most, or something else. So in a sense, I am doing the reverse of what you seem to have done and I need to reverse up a bit to better understand what 'ethertricity' actually looks like and how to observe it in a battery and then how that can be enhanced with HV only or cap dump pulse etc. I am already gathering components for a single coil system, without having to change the general physical design of my generator, (only replace the cap dump circuit with an SG one) so will get to that in due course.

      Is your welding rod strictly R60 or regular copper-coated mild steel? I can't see that a small amount of Mo in the mix would make any difference. Also did you epoxy them together or just 'ram them in tight'?

      Now I feel I ought to be doing some garden weeding!



      Click image for larger version

Name:	Pulse combining 1.jpg
Views:	350
Size:	794.6 KB
ID:	74223

      Click image for larger version

Name:	Pulse combining 2.jpg
Views:	339
Size:	633.3 KB
ID:	74224
      'Consciousness came First'


      • #93
        Hi Julian,

        Yes, will do more proper measurements, probably as you described them, once I;m getting closer to the right settings for the battery I’m currently using.

        HV peak measurements
        The attachment I uploaded contained a mistake: 3rd test was not done with STP20N95K5 but with STP12N120K5. I corrected it and uploaded the document again as a REV 02.
        I did not use the voltage divider, but my 100:1 probe. I did not record the temperature, will try to do so in the next test, but they were both done around 10.30pm after a sunny day, so temperature should have been more or less the same in my room.

        You state: “so far I have not found the charging battery to rise above its starting voltage when swapping is involved and the elevated charged status is then used as the supply.”
        That statement applies then for the pure radiant charging correct?
        As mentioned before; JB said that this would not work by swapping the batteries with just radiant charging. However, I do not say that we should not retest and try it anyway and/or in different setups that JB has not tested yet. And when thinking about his statement, I wandered how it applied to his 1984 machine since there he was using 1 battery part of the cycle cranking the shaft, part of the cycle catching the ‘radiant’ spike from the coils…
        If I remember correctly Paul Babcock mentioned that the timing of the swapper was very important, around 3min out of the top of my head.
        As mentioned before, some other routes to try with the radiant charging and the battery swapper is the setup mentioned in the 3rd bedini handbook (out of the top of my head), where there are 3 or 4 batteries involved (I understand changing to that setup might not be done overnight). Maybe there are a couple of different setups, but I need to think about it longer… will add it on my ‘to-reply-list’ to you that is growing .
        For sure would be nice if you’d be getting a reply from Paul.

        SG setup
        Could you elaborate a bit more what you mean with going to a more SG style setup? Because if you mean using the common ground (generator mode) circuit setup, that would definitely be another thing to try with sapping. JB said that with a cap dump or common ground mode setup you can swap batteries.

        Welding Rods R45
        I did not ‘jam’ them, but they’re not epoxied in either: With a magnet I can get them out.: I still want to be able to replace them with the ferrite cores. I got 4 turned to the right size, and then my last two were to brittle and broke during turning. Ordered 3 news ones, received them, just need to take them back to the turner again.

        Best regards,


        • #94
          Hi Rodolphe,

          You have explained the different spike voltage which all makes sense. There will be some temp variation but a few hundred volts is hard to explain based on temp alone.

          Re swapping, I have used 3 min and 10min swapping with both HV spikes and cap dump discharges. I'm aware of the idea that swapping to a 'blue' battery (radiantly charged) for a supply can damage the circuit, which is part of the reason why JB proposed the cap dump system, as that converts the spikes to conventional charge.

          By SG I mean returning to the original patent-based monopole design and seeing how any radiant effects manifest before then working back up by modifying one thing at a time. I can use my rotor system by adapting it for this and then put the 'SG' circuit where my Cap Dump PCB currently is. I attach the block diagram, schematic and draft PCB for this. The present PCB v4 that you are using can be used just to manage the supply from a PSU, so standing in for a supply battery. There are some features on the PCB, such as various sockets and diode placements, that will allow various transistors and diodes to be tried.

          I see you are using R45 rods in place of the suggested R60. I can't see the small alloy variances making much difference. I won't be gluing in place and the glue doesn't add any electrical properties anyway.

          Yes, ferrite rods are very brittle!


          Click image for larger version  Name:	Bedini 1 Block Diagram.jpg Views:	0 Size:	160.2 KB ID:	74228

          Click image for larger version  Name:	Bedini Schematic 1.jpg Views:	0 Size:	526.3 KB ID:	74229

          Click image for larger version  Name:	Bedini PCB 2.jpg Views:	0 Size:	267.4 KB ID:	74230

          Last edited by JulesP; 06-06-2023, 12:34 AM.
          'Consciousness came First'


          • #95

            Hi Julian,
            Had some time to kill on a train ride yesterday and made the most out of it :

            Post #83 - max charging voltage with regular charger:
            The voltage where the charging curve became horizontal kept on rising with every charge/discharge cycle I did with the regular charger. Once it hit +/-14.26V the battery charger did stop charging (instead of continuing after 10+hours).

            Post #90 – Slow switch off of active device
            You said “Your spike voltages are good for the device used (IRF840) although seem to be switching off slower, hence the curved slopes.” When you say “slower” do you mean slower than the spike rises (switches on)? Or did you mean slower than that your IRF840 switches off? If the latter is the case, any idea what might be the cause of this? In the measurements my coils had less inductance than you, so if I expect to see anything, I’d expect them to switch off quicker (unless the inductance is not the cause of the slow switch off). My measurements on the scope for the STP20N95K5 and STP12N120K5 look the same, so the IRF840 itself was not the cause.

            Post #90 – Amp meter
            “PS. Please note that my inserting the ammeter into the negative line via the plug panel is not in fact the best place for it to get an accurate reading. I found that going from the supply battery directly to the meter and then out to the negative line was best. I can give more info on that if you need. Of course, with battery swapping, you can't really use a digital meter since when the meter is in battery line during the charging stage, it won't like the spikes and may damage it so I used the configuration shown with swapping off for the CoP tests, and also similarly when substituting the supply battery with a PSU.”
            The panel meters I have installed as show on the diagram (so in the central negative line), but the digital multi meter (DMM) (that I finally started to use ), I connect directly to in the positive line from the PSU, and then from the DMM to the machine. Any particular reason why you choose the negative line? On my Bedini SG I experimented with a fairly accurate Analog meter in the negative and the positive line, but for me it did not make a difference.

            Post #92 – Paul B Voltages
            “It is possible then that at the voltages that Babcock uses to go beyond the 'ether tidal region' as he describes it.”
            Have you tried to go to these voltages? (don’t know out of the top of my head what they were)

            Post #92 – Impedance
            “…impedance is the major blocking factor and where it matters most”
            Also here; With my SG I went pretty extreme with replacing all cables to something like AWG 10 or 8, but for me it made no difference at all. It could be however, that it only starts to make difference if your machine is producing some over unity already, in other words; when your pulsing at the right frequency… My very limited adjustability of the frequency/RPM of my SG might have been the first bottle neck so I could not notice (yet) the difference that the thicker cables made.

            Post #94 – SG setup
            There are still a couple of things that are unclear to me regarding your next step going to a SG setup:
            -“..monopole design and seeing how any radiant effects manifest before…” Which things would you be able to see/study with the SG setup, that you cannot with your current Solid State design (SSD)? The whole reason for me that I moved to your SSD is that I did not have enough control over certain parameters, that the SG has not, or at least not in an easy way…
            -“ The present PCB v4 that you are using can be used just to manage the supply from a PSU, so standing in for a supply battery.” Why would you need your PCB for this? -> Why would you not be able to connect it directly to the input on your proposed new PCB?
            -In case it helps -> I probably still have an unused 8 transistor PCB board from TeslaGenX for the SG lying around here… But with the customization you want to do, it is probably of no help.
            -Some suggestions/additions for your prototype new PCB:
            --have the option of adding more (Transistors) coils in parallel
            --have the option of adding more D1 diodes in series
            --have the option to switch to Common ground/generator mode
            --Maybe you want to swap out your diodes for different version (UF / or silicon carbide versions) and make a feature for that. But if you don’t mind soldering them in/out, that it is not necessary of course.

            Progress on my tests
            I did runs @ 100Hz, 110Hz, 120Hz, 130Hz, 140Hz, 150Hz, 160Hz, but did not notice a mention worthy change in COP. Will continue up till 200Hz, and then scan the a bit in the lower ranges (90Hz 80Hz 70Hz 60Hz, after that continue above 200Hz, or go in the steps between; 105Hz, 115Hz, 125Hz, 135Hz, etc.
            All done so far at 50%duty cycle, welding rods as core, majority done with STP20N95K5.

            Best regards,


            • #96
              Hi Rodolphe,

              Re 90: Your spikes seemed to have a distinct curve at the bottom which may be due to capacitance of some sort but also, as I suggested, due to the time base setting that spaces the spikes out. I attach one of mine for comparison. Any difference may have no significance but it could be due to my using ferrite cores and not R45. Stronger mag field and probably less hysteresis so sharper cut off.

              Re 90: The reason my design has a common earth is because of the swapper so that the positive line can be swapped as the supply without having to also flip the negative. With a common ground line the obvious place for the main switch, fuses, and meters is in that line otherwise you would have to change the contacts for measurement every time the swapper operated. In terms of measurements, having the meters in either line makes no difference, so I could have had a common positive and switched the negative. However, since many other parts of the circuit need to link to a common line and it is 'tradition' to link the negative ground lines together, then that is the route I have taken. When it comes to measuring the current, since I was staying with Battery 1/PSU as the supply (so swapper off), then for the measurements I could just insert it as close as possible to the source in the distinct line i.e. the positive. Also, since I would often get slightly larger readings when my meter was placed near the main switch and fuses, I got the impression that it was not the best place to put a meter (ok for the inaccurate panel meters) so for accurate measurements, like in CoP tests, I opted for the positive line close to the supply as you are doing.

              Re 92: I have tried a 4.5kV device but I need to modify my divider to a 20:1 arrangement as the limit on my scope is 400V. Also, since my scope has died after 5 years and another is soon to arrive, I will soon be able to measure the spike output with this IGBT (IXYH30N450HV - expensive but the best value was Mouser). An added complication is that the leg format is different from the 1.7kV device and so I have had to use some wires that cross over the solder points. Whenever I get around to updating the PCB I will include a site for the TO-247HV format as well as the other two: TO-247 and TO-220 as well as isolate the collectors/Drains so that all three devices can be installed at once without burning each other out. In due course, I will see the effect of these higher voltages as part of my larger bank of power tests, as indicated below.

              Re 94: The only reason why I was considering going 'backwards' to build a single or seven coil SG machine was in the hope of observing 'radiant effects' as per the SG-1/2/3 manuals and then cross-reference to my system to see what features are important. Now that I am getting some promising power data I no longer need to do that but am focused on adapting my system to be able to switch between my current 'common ground' configuration and the 'Classic SG' and the Genertor' modes. I have plenty of testing planned with all these modes and using the rotor, cap dump, pulse combing and straight HV, and with up to 4.5kV, or as high as my system will go with that IGBT.

              The reason I suggested using my PCBv4 as the power supply is that it is already there doing that (via the swapper relays, Buck/Boost converters etc) so there was no point in removing it and leaving a big hole and having to fill it with something doing essentially the same.

              Re your progress: If you like I can post here or send you a precise methodology for you to try that I used for my voltage-based CoP tests. Once you have a standardized method then it makes comparisons more appropriate. If you are using Ah(charge) for your measurements, instead of energy (Wh=J) then that may give quite different results. What battery and CBA are you using?

              Kind regards,


              Click image for larger version  Name:	IRF840 spikes.jpeg Views:	0 Size:	863.2 KB ID:	74242
              Last edited by JulesP; 06-14-2023, 05:12 AM.
              'Consciousness came First'


              • #97
                Hi Julian,

                Thanks for all your answers.

                To start with the last part since it is most crucial in my tests for now:

                I deviated slightly from my previous method in the sense that before I took manual readings and used for my Ah calculation my manually noted down amp reading at 60% of the charging time.
                That process I automated now by using the digital multimeter, taking automated measurements every 5min. Although I still use the 60% reading (to compare with my previous measurements) I also now have a new measurement method, by taking the average of the readings during a charge cycle. In other words:

                Old method:
                -take amp reading @ 60% of charging time
                -with the above mentioned amp reading and the total time it took to charge I calculate how much Ah is supplied by the Power unit (PSU) as input
                -With this Ah value and knowing that I discharged 1Ah originally out of the output battery, I can calculate my COP.

                New method:
                -auto record amp values every 5 min, at the end of the cycle add all the values together and divide them by the total number of recordings
                -with the above mentioned amp reading and the total time it took to charge calculate how much Ah is supplied by the Power unit (PSU) as input
                -With this Ah value and knowing that I discharged 1Ah originally out of the output battery, I can calculate my COP.

                Calculating it in J would in relative sense will not make much of a difference:
                For charging, the input the PSU delivers constant 12V, the amperage draw and time (based on my measurements) are also fairly constant. Same goes for discharging: amperage draw is constant 1A, time is constant, average voltage would only change very minimal.

                With all the measurement I’ve done so far, the difference in COP is all within 5%. My plan was to go to more accurate measurements once I see a change in COP, but since it is all in the range of 40-ish% a necessary increase of more than x2 will show up very clearly also in my inaccurate way of measuring.

                I scanned the following frequencies: 40Hz (testing while writing this), 50Hz, 60Hz, 70Hz, 80Hz, 90Hz, 100Hz, 110Hz, 120Hz, 125Hz, 130Hz, 140Hz, 150Hz, 160Hz.
                Around 100Hz it seemed a little bit higher, so will go there in the steps in between next week probably (e.g. 105Hz, 115Hz), but I don’t expect to see a huge jump suddenly.

                Other settings:
                -Duty cycle 50%
                -Active device STP20N95K5
                -Welding rods as Core material
                -12V input voltage (PSU)

                I still plan to do a quick test with only 1 coil connected, with very short connection wires and the ferrite as core material, to see if scope reading of the HV pulses will become similar to yours.

                Output battery that I use:
                VMF - 52805 12N24, 24Ah 293A(EN), Flooded lead acid.

                CBA that I use is a CBA V (West Mountain Radio), I have a IV as well.

                If you think I’m missing something, or should focus on something else first instead of the frequency, I’d be very keen to hear. Otherwise I’ll continue as mentioned: scan some frequencies around 105Hz, 115Hz, and might consider going further down, 30% (depending hoe current test turns out) and maybe further up 170Hz, 180Hz.

                After that I might try another battery.

                Best regards,

                Click image for larger version  Name:	Post 97.PNG Views:	0 Size:	27.2 KB ID:	74246
                Last edited by pearldragon; 06-17-2023, 03:26 PM.


                • #98
                  Hi Rodolphe,

                  I think that as long as you have a stable and repeatable method then that will enable you to see changes in the battery response and hone in on the optimum PRF (and other variables) for your setup. Each battery appears to have a natural ringing frequency to the pulses.

                  I certainly found that the response can change more than 150% over a 15Hz range as the attached graph indicates. Normally you find a raised value in a frequency zone and then zoom in around that frequency. You can see from the manual the PRFs I found for a variety of batteries. But again, slight deviations in the methodology will introduce errors that can mask that fine-tuning and so being consistent is important.

                  Click image for larger version  Name:	Optimum PRF.jpeg Views:	0 Size:	275.5 KB ID:	74248
                  Once you have the optimum values then you are getting the best from the battery when it comes to determining power outputs, which is the stage I’m at now.

                  Finding the power output for a cap dump and rotor is relatively straightforward (although two equally valid calculation methods give values different by a factor of 2 since one method is a dynamic one and the other a static one, so I’m investigating that) but for the receiving battery is rather more involved. Soon I will be writing a document outlining a method of deriving the battery CoP and passing around for comments.

                  As we have seen, CoP values based on elevated charging voltages do not relate to power output in a straightforward way so that stage should perhaps be conceived of as one to find the best battery response in preparation for real-time power output measurements and then related to power input values.

                  Also, I am beginning to see that the optimum charging, and for ‘radiant’ effects to manifest, is not just about the potential, as has been implied by various contributors in the past, but also requires a minimum amount of charge. That being so, your CoP measurements based on Ah (charge) may well be more indicative of available power than ones based upon potential alone.

                  Lastly, reconfiguring my system to be in the traditional SG mode is straightforward for the rotor system but not so when I add in my cap dump circuit. This is because I designed it as a high-sided one (switching the positive cap to the battery line instead of the negative) to go with my swapper. I am going to construct a low-sided version (a simpler circuit) so that I should be able to integrate it with SG mode (no swapping) and see how it performs.

                  I guess one day I will come up with an answer to my original research question and how to achieve it.

                  'Consciousness came First'


                  • #99
                    DEVELOPMENTS UPDATE

                    Here is an attached summary of recent work, developments and trends.


                    Developments Update.pdf
                    Last edited by JulesP; 07-05-2023, 03:40 AM.
                    'Consciousness came First'


                    • Hi Julian,

                      Complete misunderstanding (my side) regarding function of your machine (Classic SG versus Generator mode)
                      Reading your post/document post #99, I realized I have been misunderstanding in which operation mode your machine has been working… (If I’d spent a couple of minutes to check the schematic more in depth I would have known though.):
                      I was under the illusion that your machine (as you initially designed/uploaded it) was working in Classic SG mode (pure radiant). Now reading your post #99 I see that it is Common Ground mode (or Generator mode, in the Bedini Handbooks both terms are used).
                      This means that my comments of posts #67/#93 still hold in abstract form, but not in relation to you machine: Only classic SG mode = Pure radiant. Common ground/generator mode has also something else drawn from the input battery. I’m not sure if you can call this ‘something else’ regular current, but how I understood it from the DVDs (and Garry) is that you should be able to swap the batteries when charging in this mode.
                      Now also I finally understand whey you decided to go back and test with the Classic SG setup (pure radian). As explained, I continued to not understand what you’d learn from it since I was under the impression that your machine was running in the same mode. It was not. It all makes sense now.

                      Another reason why I was under the impression that your machine was running in classic SG mode (pure radiant) was the following:
                      When I first started testing with my Bedini SG, in Classic SG mode (pure radiant), I used the same PSU as I use now as input. When I modified my machine later to Common ground/generator mode and tried continuing using my PSU as input, it went bunkers: it made all kind of clicking noises (safety/overload switches being triggered) and the power supply was everything but constant. This was due to the fact that in this mode the PSU was exposed to some/part of the HV spikes too, which would trigger safety/overload relays/features.
                      However, when I used the same PSU as input on our machine it did NOT show this behavior at all… Why it does not, is not clear to me yet. I guess some of the components eat up that spike before it hits the PSU… but that means your system DOES perform different than Bedini’s circuit in Common ground/Generator mode, meaning that potentially your input battery does NOT gain the benefits of receiving part of the spike too… However the fact that it does not, does make it way easier to find the sweet spot with a stable PSU as input. But would be very worthwhile knowing which components in your circuit are responsible for this change in behavior.

                      Another thing that might (or might not) relate to this: In your report of post #99 you state that you’ve not been able to charge above 13V. Have you tried playing around with the Duty cycle to influence this?
                      On my Bedini SG I could indeed go up to 15+V, I don’t think I tried as high as 16V, but assume it would have been possible if the batteries chemistry would allow it.
                      With the battery were I’ve done the majority of testing with on your machine (24Ah, LA), I normally charged to 14.3V, since in that area was also where my regular charger would charge it to (after doing several cycles with it).
                      With the 12Ah AGM I charged to 14.84V, since that was the point I charged it to with my Renaissance charger from TeslaGenX before discharging 1Ah out of it. I did notice though (since I was too late to shut off your machine) that it could not push it over 14.84, but it actually declined back a bit while leaving it charging (accidentally). I’m pretty sure that if I would have charged it beyond 14.84V with the TeslaGenX charger, your machine would have charged it to that value as well. By playing with the duty cycle I probably could have pushed it further if I wanted to.

                      Update on tests:
                      As mentioned before I tested from 40Hz-160Hz, played around a bit with Duty Cycle and different active devices, but all with little effect on COP, around 42%. When factors did have effect, it was mainly influencing the COP in a negative way, but nothing that showed in improvement. A friend of mine suggested that instead of discharging 1Ah out of the battery I could also go for a lower amount (e.g. 0.5Ah) to test faster. Seemed like a good idea, but in practice this resulted (unexpected) also in a drop of COP.
                      Looking at Fig 52 out of your manual (also in your post #98) I used increments of 10Hz judging that with those steps I would see a hint if a sweet spot was getting closer. When I thought I saw a potential hint, I decreased steps to 5Hz, but all to no avail.
                      Since the battery I was testing with (see post #97) was the one that also gave the worst results with my Bedini SG, I decided to give another battery a try that gave better results with my Bedini SG: A 12Ah AGM battery: Xtreme 82-216#, 12Ah

                      When in the first run it also showed 41% I decided to stop testing and reconsider my setup: That these two different batteries gave almost the same COP, gave me the idea that there is another bottle neck in my setup at the moment.

                      Question: Looking at that Fig 52 (post #98), the bell curve shows COP 1.37/1.38 at the ends. How does the COP look in the areas further away from the peak? Is it all above 1, or did you to have the majority of the COP points in the 0.4 / 0.5 range? And thus it could be that I just did not scan far enough (that would mean above 160Hz, since I find it unlikely that the bel curve would be below 40Hz).

                      Strange and persistent Tail on HV spike:
                      One of the things I would like to find an answer to before continuing is why this “tail” shows up on my spike on the scope. You suggested that this could have to do with the (horizontal) scale on my scope. It is 5ms on my scope, and in your post #96 it is 4ms, so in the same range.
                      I tried the following to test see if it would influence this “tail”
                      -disconnected 4 of the 5 coils and connected the remaining one without the M5 bolt to eliminate potential too much capacity from the coil to the PCB (these M5 bolts connect the enameled copper wires from the coil to insulated stranded wires that go the PCB -> No effect
                      -disconnected 4 of the 5 coils and removed the welded rod core material (so air coils) -> No effect
                      -disconnected 4 of the 5 coils and filled the core with a ferrite core (coil inductance 400mH) -> No effect
                      -Reading your post #99 -> replaced the input PSU with a battery -> No effect

                      Next thing I plan to do is connect my scope again to the output signal of the FET driver and again have a look how this signal looks like with ‘HV LOAD” switch switched ON (and with another scope connected) between the output test point. Regardless how the signal looks like I might try to switch out that FET driver again and see if that makes a difference.

                      Regarding your COP test method, page 73 of the manual:
                      1. -> As showed in my previous posts, I would not do just one cycle with regular charger and take that as a reference, but do this a couple of times until the peak voltages of the charging process with the regular charger stabilizes. This applies both for new and uses batteries.
                      2. -> As mentioned before a discharge rate of 3A on a 7Ah battery is huge. Again, I have not tested with Li-Ion batteries yet, as you mentioned; they might handle it better than AGM/LA.

                      Duty Cycle:
                      The unit to set the frequency and duty cycle seems to work in reverse regarding Duty Cycle: When I increase the Duty Cycle (on the display) my amp draw goes down, while when I decrease the DC on the display my amp draw goes up. Do you experience the same?

                      Best regards,


                      • Hi Rodolphe,

                        I designed my generator to have a common earth since that allowed for a battery swapper that switched just the positive line. I have recently modified the V4 circuitry to allow switching between common earth, classic SG, and generator modes and where the generator and common earth modes are the same except for the inclusion of an extra diode in the former.

                        A good diagram to see the earlier 'common earth' only wiring is on the front cover of the manual. The common earth connection point is clearly visible at the top left.

                        The figure below shows the recently revised v4 wiring to be used with the new Low Sided Cap dump circuit that I am setting up and testing now. Notice the change to a 'non-combined' earth post for the mode changes, compared to the original V4 design.

                        Click image for larger version  Name:	Revised Wiring for SG-Gen with LS-CD.jpg Views:	0 Size:	361.9 KB ID:	74270

                        You refer to SG mode as radiant mode but so far the only advantage I have seen in using that mode is the reduced supply current, down to about 20% of that in common earth mode.

                        This is certainly helpful in allowing any net charging to be more visible compared to the drain on the supply battery (see Fig 4 in my latest update report and below), but so far it has not resulted in more radiant energy effects being observed.

                        Click image for larger version  Name:	image_6759.jpg Views:	6 Size:	309.3 KB ID:	74267

                        With your earlier SG build you seemed to be getting some good charging but, if it were performing as Bedini et al. proposed, then surely you would have got a clearer CoP>1.

                        This adds weight to my plan to use the original SG circuit to find out where and why many SG builds have not been as successful as hoped. As I say in the report, one can’t expect to build a more powerful version if there is a fundamental ‘pinch’ point or blockage in the basic build. All we do is complicate the problem and make it harder to achieve our goals.

                        So not having 'Classic SG' wiring in my original design is only part of the reason why I need to take a temporary lateral step.

                        I hear what you say about my common earth mode being somehow different than generator mode, regarding the use with a PSU. Looking at the classic SG and generator modes on the right of the first diagram above, I’m not clear how generator mode results in HV pulses impacting the supply battery. My common earth pathway is effectively the same as the Generator one but without the extra DG diode. With reference to the ‘Generator mode circuit’ diagram on the right, the HV pulses meet the capacitor and are absorbed there. Possibly some could reach the receiving battery directly, but there is no obvious pathway for them to find their way to either pole of the supply battery on the left. That suggests that in your SG build there was some ‘accidental’ pathway where they could do that. I certainly can’t see one in my V4 build which is why you don’t get that interference. If JB’s design and intention were to have the HV spikes affect the supply battery as well in Generator mode, then please indicate the proposed pathway.

                        When it comes to SG mode, then I can see how the impact of HV spikes on the receiving battery can also affect the supply battery. Even if C and S were not in the circuit (i.e. no cap dump circuit), the two batteries are effectively in series and so some charging effects will be carried through to both when transistor T closes.

                        I have not been able to take a battery above approximately 13V with either HV or cap dump charging. The over potentialisation of the electrolyte is a feature of radiant charging and the fact that others have been able to hold voltages for a period of time over 12.7V indicates that other processes are at work that go beyond the ‘normal’ thermodynamics . The electrochemistry is such that under normal charging it should settle back to around 12.7V when fully charged, even if charged at 14.2V to drive the process forward in a normal charger.

                        You say you have reached 14.84V with my design which is very interesting. Can you clarify the FET used, PRF, duty, charging time, and supply current? Presumably, no swapping was used there. Do you have any voltage-time plots you can share, with or without swapping; like mine here? This one shows the result of 'punctuated' swapping where after each swap a rest period is allowed to allow the voltages the settle. if even a moderate degree of radiant effects were occurring then we should expect to see a significant and stable voltage rise - not a gradual net decline.

                        Click image for larger version  Name:	Punctuated swapping with HV pulses.jpg Views:	0 Size:	280.0 KB ID:	74264

                        I usually find that the charging rises quite fast but levels off around 13V so there’s an issue to explore. I haven’t used a Renaissance charger or similar but I assume these are HV pulse systems?

                        When you got voltages of 14.8V with your SG and the v4 generator, do they remain there or, over an hour or so, fall back down to perhaps only a fraction of a volt above the starting value, as mine do?

                        The high values of CoP I derived were based on the voltage changes as described in the manual. I have recently heard of others who have observed this so-called ‘phantom voltage’ effect. This is where increased voltages in the receiving battery vanish when the battery is put under load. It’s as if the voltage increase is not secured into the electrochemistry but perhaps lurks as a surface charge on the plates.

                        Regarding the CoP graph, I didn’t go beyond the tails shown in that case. Once it was clear that the CoP was rising, then I focused on that area. Besides I have never achieved a CoP (based on voltages) lower than 1.7 with HV pulse charging. However, as I am finding out, this does not translate to an equivalent CoP in terms of power, possibly due to a lingering 'Surface charge effect - the effect that causes the substantial initial voltage rise during charging.

                        Since you mention experimenting with more or less coils, here is a graph showing the use of 1 - 5 coils.

                        Click image for larger version  Name:	CoP vs Number of Coils.jpg Views:	0 Size:	122.7 KB ID:	74265

                        It is strange that you get a spike ‘tail’ compared to mine. I think doing what you are doing is the only way to define out why. One can bypass the driver by taking the chip out and inserting a small jumper lead between pins 2 and 7 of the dip socket. This takes the input signal straight to the output pin. The jumper lead is made with two small legs extracted from a socket and some fine wire (see pic). The driver output should be a fairly clean square wave.

                        Click image for larger version  Name:	Wire jumper.jpg Views:	0 Size:	537.8 KB ID:	74266

                        Re page 74 of the manual (not 73) discharging at a rate of 3A (0.42C) is not extreme for a 7Ah battery, especially for just the 20mins required to drop 1Ah. The battery is then left 1 hour to settle before any charging. Try 1A for an hour for comparison, but heat is the issue and the battery spec sheet tells you the maximum safe discharge rate. With my battery (LP12 -7) the maximum discharge current for 20mins is 8-9A so I’m well below that.

                        Lithium batteries do respond well to radiant charging but it will be far more interesting to compare performance when I get to see some actual radiant effects. The charging gradient is shallower than with LAs so for a given energy input the voltage rise is less.

                        Yes, if I increase the PWM duty cycle the supply current goes down, which is the opposite of what I would expect. I settled on 65% for my tests. I have not yet got my hear around why the supply current behaves this way. It's as if the circuit has to acquire a certain amount of energy for the coils to do what they need and if you give it less time to do that, by reducing the duty cycle, then it has to up the current to achieve its needs. Sounds almost conscious!

                        When using the cap dump system then I found that the optimum cap charging occurred at a PRF of 54Hz and 40% duty and where using 1.7kV over 600V gives a modest improvement of about 20% and not nearly 3 times better, as one might expect.


                        Last edited by JulesP; 07-10-2023, 01:11 AM.
                        'Consciousness came First'


                        • Hi Julian,

                          I’ll go in through your response in a bit of a chaotic order. The reason that I do this is that I see that we’re starting to misunderstand each other regarding some definitions and interpretations and I want to address those items first hoping to get our thought trains back on track there

                          Common ground mode & Generator mode:
                          You’re defining the two as different, the difference being an extra diode. As I understood from Gary (and I agree) this extra diode is only there to prevent the input and output battery from (partially) equalizing instantly when the circuit is fully connected when the input battery would sit >0.6V in resting voltage higher than the output battery, assuming a diode with a Fv of 0.6V. If the input battery would sit >1.2V higher than the output battery, you’d even need another extra diode. If your input battery resting voltage is <0.6 than the output battery, you can leave it out.
                          The terms Common ground mode & Generator mode come from page 21 of the Advanced Bedini SG manual, are used interchangeable there and so I’ve used them interchangeable as well: there is no difference in operation principle.

                          Common ground (CGM) mode & Generator mode (GM) impact on PSU
                          “I hear what you say about my common earth mode being somehow different than generator mode, regarding the use with a PSU.”
                          My point was not that your Common Ground mode has a different effect on the PSU than Generator mode. As described above, they’re the same, that diode does not change its workings. What I intended to convey, is that although your circuit is in theory as much an CGM/GM as the circuitry in from the advanced Bedini SG handbook is, the same PSU that I used for the Bedini SG in CGM/GM was unusable with the Bedini SG CGM/GM, and can be used for your circuit. In other words, some of the component (maybe from the auxiliary circuits) in your circuit change the impact on the input (PSU), (and so on the input (battery or PSU)).

                          “If JB’s design and intention were to have the HV spikes affect the supply battery as well in Generator mode, then please indicate the proposed pathway.”
                          In my responses below I leave cap dump circuit out of the equation for no since I did not experiment enough with it so say something sensible about it.
                          In the two images on the left side, Fig 3 of your last posted document “DEVELOPMENTS UPDATE”, there you see that the input (battery or PSU) is included in the circuit when the HV spike is discharged in CGM/GM.
                          Although you’ve indicated it almost the same already in Fig 3 in your update, in attachment 1 another way of visualizing it. What JB intended (in my opinion), since the input is in the circuit when the HV spike discharges, it draws current from the input in its wake. So what hits the output battery, is a HV spike in combination with some normal current. (although I’m not sure if ‘normal’ here is the correct term).
                          You might not expect it, but BOTH in Classic SG mode AND CGM/GM the input (battery or PSU) receive some sort of pulse from the coil discharge. If you measure at your input battery you’ll see it on your scope.

                          “When it comes to SG mode, then I can see how the impact of HV spikes on the receiving battery can also affect the supply battery. Even if C and S were not in the circuit (i.e. no cap dump circuit), the two batteries are effectively in series and so some charging effects will be carried through to both when transistor T closes.”
                          But what happens is opposite to what you describe here: it is when the Transistor OPENS (=not conducting) when the spike takes place. With the transistor open (=not conductin), the input (battery or PSU) is at one terminal disconnected; where it connects to the transistor -> emitter side. But, as mentioned above, if you measure around the terminals of the input battery, although it is only connected with 1 terminal during the HV spike discharge you WILL see a spike there as well…
                          (in case you’re interested I do have somewhere a CBA recording of my Bedini SG in CGM/GM where I completely disconnected 1 of the terminals to the input battery, connected my renaissance charger to the output battery, and still would see that it would charge the input battery AS WELL!!)

                          Charging profiles
                          Please find the data you requested in Attachment 2 (24Ah FLA) & Attachment 3 (12Ah AGM). Just to clarify: these high voltages I mentioned in my previous post are at the peak of the charging cycle, NOT resting voltages.

                          Classic SG mode 20% less energy draw
                          “You refer to SG mode as radiant mode but so far the only advantage I have seen in using that mode is the reduced supply current, down to about 20% of that in common earth mode.”
                          According to the advanced SG handbook the CGM/GM uses indeed a bit more energy but the output is more as well resulting in a net higher COP. That is according to Lindemann’s setup.

                          Bedini SG
                          “With your earlier SG build you seemed to be getting some good charging but, if it were performing as Bedini et al. proposed, then surely you would have got a clearer CoP>1”
                          “This adds weight to my plan to use the original SG circuit to find out where and why many SG builds have not been as successful as hoped. As I say in the report, one can’t expect to build a more powerful version if there is a fundamental ‘pinch’ point or blockage in the basic build. All we do is complicate the problem and make it harder to achieve our goals.”
                          See my last post #206 in the thread where I discussed/compared/upgraded my Bedini SG as much as possible to match Gary’s results:

                          Renaissance battery Garger TeslaGenX
                          This is HV spike cap dump charger. So like you’ve been experiencing with. Also discussed in the Intermediate handbook, with the difference that it is a Solid State version (like yours).

                          Testing with 1 & 5 coils
                          “Since you mention experimenting with more or less coils, here is a graph showing the use of 1 - 5 coils.”
                          I only did full charge/discharge cycles with 5 coils.
                          Only to test individual parameters I sometimes connect just 1 coil to make sure I isolate the parameter and don’t have any interference from multiple coils.

                          Best regards,

                          Post 102 - Attachment 1.pdf
                          Post 102 - Attachment 2.pdf
                          Post 102 - Attachment 3.pdf


                          • Hi Rodolphe,

                            It’s been a while since your post but, with so much going on at the moment, I will only be able to reply once a week.

                            Wiring modes:

                            I think our wiring depictions agree. My main point is that, despite the ‘Generator’ mode being considered to give better results, I get the best results so far in SG mode, primarily due to the reduction in the supply current. Hopefully, that will change in the future.

                            At the moment I can’t explain why my ‘Common’ mode is not affecting your PSU (or supply battery when used) in the way it was affected in your SG build. I expect some explanation will pop into view at some point. However, I see the fact that my design doesn’t interfere with your PSU as further evidence that it is not producing full radiant effects; so as a negative. I would be more encouraged if it did upset a power supply, which might be expected if ‘radiant’ energy was more evident. So it’s not that the V4 circuit has or has not got one or other components, or that its Common Earth mode is somehow significant, but that, as a system, its radiant effects are minimal.

                            The spikes will impact the supply battery in SG, Gen and CGM modes since the battery, or the positive terminal in SG mode, is connected to the spike pathway in some or other fashion. How it impacts it is not known or clear at the moment.

                            In the same way, different batteries result in different current draws, even if their voltages are the same. The micro energetics are clearly complicated. In my earlier text, I meant to say T opens, as the spike appears when the transistor has just opened, as per regular electrical theory.

                            Charging Profiles:

                            I am particularly interested in the charging profiles you have obtained with the v4 board. After several hours you have reached approaching 15V, even if it does drop back again a lot. I have so far only achieved approaching 13V, but then I have not regularly charged a battery for 3 hours continuously.

                            I wonder if the different type of terminals used by us to connect the coils makes any difference?

                            I have just started to see how an SG circuit performs using one of my present coils (14 Ohm). I have not yet wound a new one incorporating a trigger winding so, while I’m setting up to do that, I’m running the BD1 board using the 2N3055 transistor, IN4007 output diode, in SG or Gen mode and using the PWM module in lieu of the trigger coil. The board can be switched to use the MJL21194 transistor, an Silicon Carbide output diode and, as mentioned, run with a trigger coil.

                            First observations are positive in that the residual voltage increase after 30min charging is significantly higher than I have seen using my system. Also the raised voltage is more 'solid' and not prone to vanishing the moment a load is put on. Compare a 3 hour charging curve I did yesterday with the BD1 circuit, with one of yours. A very different shape and yours is more in line with that presented in the SG-2 book (see third graph). Your one fits very well with that, ignoring the rapid fall off after turning off. If only it was clear why ours are so different, that would solve a large part of the problem.

                            Your battery is an AGM and not fluid filled, yes? And what was the resistance? Still 14 Ohms?

                            Click image for larger version  Name:	BD1 Charging Curve.jpg Views:	0 Size:	223.7 KB ID:	74277

                            Click image for larger version  Name:	Rodolphe curve.jpg Views:	0 Size:	115.0 KB ID:	74278

                            Click image for larger version  Name:	Desired charging profile.jpg Views:	0 Size:	132.0 KB ID:	74279

                            I have just measured the peak spike voltage at 130V and it is possible that the type of active device is of much greater significance than previously thought. That and the coil impedance, and where a 4 Ohm coil will probably perform much better than my 14 Ohm one.

                            I don’t know much about the other chargers you refer to. Do the Teslagenx chargers result in a CoP>1, if they are based on Bedini’s original circuits? If not, why not?

                            From memory Gary told me his best CoP was around 1.05-1.1.

                            I will pass on any useful ideas and data as they emerge, but on a weekly basis or thereabouts.


                            Last edited by JulesP; 07-20-2023, 02:00 AM.
                            'Consciousness came First'


                            • Hi Julian and Rodolphe,

                              Rodolphe asked me to respond to his observations of the radiant affect on the PSU as discussed in the last two posts of this thread. I did that in a PM to Rodolphe. But I would also recommend a download from E-Media Press for each of you to purchase, which helps explain what's happening in these circuits. It's a $17 package of two videos (one by Arron and one by Peter) as well as a simple PDF file drawing of the circuit under discussion in the videos.

                              Gary Hammond,


                              • Hi Gary,

                                I would imagine that, apart from showing actual devices, the videos have more info than the SG1-3 books? I’m sure they will be of interest. I’m also liaising with Peter on the issues.

                                I have installed an ‘SG’ circuit board and will be able to wind the dedicated coil soon.


                                Last edited by JulesP; 07-24-2023, 09:11 PM.
                                'Consciousness came First'