An energy harvesting project

[pearldragon]

Hi Julian,

I’m watching some youtube video’s about mosfets (switching). In the following YT movie @52 sec, it shows the mosfets with an already built in diode to protect against/ground the spike during the switching on of the mosfet. Which is further shown in the Vds and Vgs Graphs @3min: only a spike during switch off.
If I open the specsheet of one of the mosfets you named in your documents/we’ve been discussing here, I also see that diode: https://www.st.com/resource/en/datas...tw12n170k5.pdf

So my question is:
Is this built-in diode not already doing the grounding job, where you intended the DSEI-12-12A/DHG10i1800PA for?

Best regards,
Rodolphe

[pearldragon]

Hi Julian,

I ordered the 1800V diodes from the Aliexpress link you posted, thanks: I figured for the money they cost there I shouldn't doubt and just buy them.

I did a RFQ for the 1800V diodes and the 1700V & 1500V mosfets at jhykjic.com (from proper manufacturers). I'll see what they come back with. Jinftry could supply the 1500V mosfets, but not the 1700V versions. If jhykjic.com can't either, I'll just go for the1500V versions.

I ordered the (semi) automatic coil winder from my post #11.

Thanks for all the quick replies.

Best regards,
Rodolphe

[JulesP]

You don’t need a high current rating and the DSEI12 will not work with the 150 or 170 as it will breakdown at a lower voltage than those with the spike we want, on FET shutoff.


I have never used two diodes in series so don’t know how that would work. If that diode breaks down at 1200V max then I’m not sure that two in series will manage 2400V. As you say the PCB is set for one device at that point.

The Ali express ones I got are fine and work well. Yes it’s possible to get duff devices but these are good ones. Visit the following link on AliExpress: https://a.aliexpress.com/_EI3y3L7

The dodgy device I had was a comparator from a UK source.

When I post the component list for the ‘replication’ PCB I will indicate in some cases exactly where I got them from. In many cases, like with resistors, they should be ok from just about any source.

Jules

[pearldragon]

Hi Julian,

The DHG1011800PA diodes were easily got from Aliexpress for about 7 Pounds for 10pcs with delivery. For the FETs try the attached.

I had a look already there and saw that they were available, but thought that buying them from Aliexpress would be risking buying 'dodgy' components as you warned me for in the cap dump thread and better left as a last option. The screenshot/link also seems a Chinese company. What was your consideration buying them there rather than from jinftry? (I just had another look at the reply from Jinftry and realize/see now they also are a chinese company).

I saw that jinftry cannot not supply the DHG1011800PA, but can supply the DSEI 12-12A. In case I want to use the STW12N150K5 or STW12N170K5: instead of a DHG1011800PA, I assume I could also put two DSEI 12-12A in series? or is there a particular reason why I should not (apart from the fact that your PCB probably does not accommodate for it).

P.S. I just see that I have some C4D02120A* diodes here already, could I use them instead of the DSEI 12-12A? I see that the DSEI have a higher current rating, but unsure if that high a current rating is necessary if it is just to ground the voltage peak during switching on of the mosfet?

*Spec sheet C4D02120A: https://cdn-reichelt.de/documents/da.../C4D02120A.pdf
Spec sheet DSEI 12-12A for comparison: https://cdn-reichelt.de/documents/da...12-12A_IXY.pdf

Best Regards,
Rodolphe

[JulesP]

Hi Ron,

This is probably not the right thread for this discussion but while you’re here I will share what I can.

I’m not aware of any process whereby water is heated first before some process maintains the heat flow. Given water’s large specific heat capacity (due to its ‘peculiar’ bonding) there must be a substantial amount of energy pit in first to reach that state.

The only valid process I’m aware of for using water as a fuel is that developed by Walt Jenkins in Florida whereby high voltages are applied to water in a way that results in vacuum energy entering the system via, rather than from, the water. This has yet to prove itself to the scientific community but he has acquired patents for it all over the world so it must have some technical merit to jump those hurdles.

For a physicists treatment of this there is Moray King’s book: Water as Fuel’ that looks at the science and then there is Walt’s site at:
https://cyberstreme.wixsite.com/mysite
which gives a fair amount of info.

Before Covid they had planned to run a bus right across America using just water but that was paused due to the pandemic and I haven’t heard yet whether that is going to be undertaken soon.

Sorry not to be more helpful re your enquiry, but at least I’m referring to the same substance!

Jules

[Ron Hammar]

Hi all, I have a need for help. I was reading on one of the forum were there was a link of a video of a patent about using water as a fuel by heating it up, The USPTO #3,939,814. Can any one recall reading about it? It dealt with heating up water over 1000*F and it caused it to cause the H2O to separate and the H2 burned off so the water being injected into the chamber keep the temp high so the action would keep on heating. I think the posting was some time ago. Thank you for any help in this. Ron

[JulesP]

The DHG1011800PA diodes were easily got from Aliexpress for about 7 Pounds for 10pcs with delivery. For the FETs try the attached.

[pearldragon]

I would be ok with uploading a parts list for the forthcoming PCB within the next week if anyone wants a head start on that issue.

You can count me amongst the 'anyone's' . I'm guessing that only the diodes and the mosfets were a bit harder to get and the other components on the list are available from the at least one of the regular component suppliers (farnell, RS, mouser, etc)? So I could place an order with Jinftry this week already just for the the diode and mosfet?

Regards,
Rodolphe

[JulesP]

Hi Rodolphe,

Yes, I have now tried the120k5, 150k5 and 170k5 with the 1800 diode. There is small improvement on CoP between the 95k5 and the120 and a bigger change between the 120 and the 150 but not so noticeable between the 150 and 170. Some battery formats (Ah and chemistry type) have a bigger improvement than others. Best CoP so far with an 18Ah LiFePO4 battery is 53.2 but I can’t say what that translates to in continuous Watts until I do power tests at the start of the year.

From my few tests so far with the supercapacitors, it appears that the battery chemistry is central to the phenomenon. That might mean that the pulses affect the chemical bonds in a way that draws vacuum energy in, just as with cavitation in water which is considered by some to draw in ZPE for the reentrant jets. This is just my speculation re the chemical bonds and I will need to consult with those who are working on that particular front.

For the best overall performance, without a big increase in Rds (Drain-Source resistance), the 150 works well but my new board will allow two FETs to be installed and then jumpers to select which one fires.

I would be ok with uploading a parts list for the forthcoming PCB within the next week if anyone wants a head start on that issue. It’s going to be 99% correct. The other components for the rest of the device will come in December.

J

[pearldragon]

Hi Julian,

In regards to #12

So far using 1.5kV instead of 1.04 and 1.1kV is beneficial in most of the tests.

Is this your conclusion based on testing with the STW12N150K5 Mosfet in combination with DHG10i1800PA diode?
(waiting at the moment for a second RFQ for the diodes and Mosfets)

In reagrds to #13

Also, I have to write an extensive 'assembly manual' explaining everything about it, together with a parts list, and I hope to have that, and the board checked, by mid-December.

Ok, great to hear! In the next on my list was having a look at the other components in your circuit, and for sure I would have questions about the exact parts you used, but I will practice a bit of patience then, if you plan to upload a parts list .

Best regards Rodolphe

[JulesP]

Hi all,

As some of you are keen to have a go at a build and have been asking about the main components etc., I attach the wiring diagram for the 'replication' circuit and have uploaded it to the 'Circuits' folder on the same Mega link. If anyone new to this thread needs the link again let me know and I will post it.

As discussed elsewhere, this is an updated PCB to the one I have been using for all my tests and is based on my findings of what works and what doesn't as, in my opinion, is not worth the effort and cost including parts that contribute little or nothing to the performance. This significantly simplifies the design, build, and components required.

The new PCB has now been designed and is off at https://jlcpcb.com/ for printing at the moment. I will assemble this board and replace my current one to check that everything works before I release it for others to use as Gerber files. Also, I have to write an extensive 'assembly manual' explaining everything about it, together with a parts list, and I hope to have that, and the board checked, by mid-December. Meanwhile, some of you may find the attached circuit helpful to use with what you already have or are compiling by other means.

A fascinating trip around the facility where these PCBs are made can be seen here: https://www.youtube.com/watch?v=ljOoGyCso8s&t=1423s

[JulesP]

I might try and measure the peak voltage and rise time with no ferrite cores and see what happens. So far using 1.5kV instead of 1.04 and 1.1kV is beneficial in most of the tests.


Two links for the type of meter are:

https://www.ebay.co.uk/itm/403872715...SMAAOSwA5di8iB


https://www.aliexpress.com/item/1005...uctlist.0.0.11

[pearldragon]

Hi Julian,

I just ordered the copper for 5 coils: 2 off -> 7.5lb AWG 21. Pricy stuff...

I am not measuring the saturation of the coils with different currents and so have no figures for how saturated the coil is.

That was what I meant/assuming with my post. That with the third dot you're more or less right at the point of coil saturation.

Regarding the diode, it seems I swapped around.

I guesses something like that happened, so I tried to google on just parts of it, but couldn't find it either

Regarding the coils, the ferrite rods are 16.4mm in diameter. They were ordered in 2018 on eBay but the order history does not go that far so I will look around when I have the time

Thanks for the diameter, don't worry about the other specs/link if you don't have them handy: I'll probably be using welding rods for the core.

Each coil took me about 10mins to wind by hand.

Thanks for that info! then That machine is NOT worth the money haha.
I'm now either considering something simple like this:

https://nl.aliexpress.com/item/32819...yAdapt=glo2nld

Or otherwise I'll go for some manual winder like you used...

I have no plans to use air coils since that will reduce the magnetic flux in the coils. But since I can argue that the voltage at the FET drain is limited by the FEt itself then you could probably do away with the ferrite cores and have the same voltage pulses. Worth a try.

I measured the combined inductance with a type of multimeter that is used for checking transistors etc. I would not worry about the value based on what I wrote above. The FET will nearly always be the factor that determines your peak voltage.

It will reduce the magnetic flux, and your FET might be the limiting factor for the peak voltage, but the dV/dt will go up because of the greatly reduced inductance, assuming that the rest of the circuit can handle the speed. But this is just me guessing and repeating information from the presentation of Paul Babcock*, which I guess Gary already has sent the link from to you.
Out of my own interest; could you tell me which multimeter it is that you used for this inductance measurement; would like to have a look at it.

*https://emediapress.com/shop/they-ar...e-vs-air-core/

Best Regards,
Rodolphe

[JulesP]

Hi Rodolphe,

The graph shows the measured CoP when the voltage across the coils was adjusted while keeping the PRF etc the same. The reason for the fall off either side may be due to several factors, one of which is that the higher voltage results in more current in the coils and so a greater energy input and so the CoP value will be pulled down. You may wonder why increasing the voltage and therefore the magnetic field in the coils does not have a bigger effect on the battery when the pulses reach it. That is so far unknown as I don't know the energetic pathways involved. It is not clear if there is a quality of the pulses that is not just expressed as its peak voltage (>1kV) but also the amount of charge it can induce (in addition to whatever it may be doing to the vacuum possibly). So there is probably another dimension to the pulse that we can't see on just a V/t graph. I am not measuring the saturation of the coils with different currents and so have no figures for how saturated the coil is. But making measurements of CoP at different voltages can make a noticeable difference to the results. Also, the optimum coil voltage is a function of the battery too in that for some 13.0V gave a better reading than 12.5. It's another variable to play with which is why I have included connections to a Buck and a Boost Converter in my circuit and the revised one to come.


Regarding the diode, it seems I swapped around the 2nd and 3rd letters. It should be DHG and I attach the spec sheet. Sorry for that; I have corrected it in the doc and uploaded it to the files.

Regarding the coils, the ferrite rods are 16.4mm in diameter. They were ordered in 2018 on eBay but the order history does not go that far so I will look around when I have the time.

I have no plans to use air coils since that will reduce the magnetic flux in the coils. But since I can argue that the voltage at the FET drain is limited by the FEt itself then you could probably do away with the ferrite cores and have the same voltage pulses. Worth a try.

I measured the combined inductance with a type of multimeter that is used for checking transistors etc. I would not worry about the value based on what I wrote above. The FET will nearly always be the factor that determines your peak voltage.

Each coil took me about 10mins to wind by hand. As it gets thicker then more wire is wound on per revolution. If you think it's worth the money go ahead but I was happy to turn a handle for a while. I did all five coils in one day.

I made the PCB using https://easyeda.com/ so it's all online and you don't need to download a client. I have learned how to do them over the past few years and there is a forum and plenty of documentation to help.

Julian

[pearldragon]

Hi Julian,

I’ve read through your Interim report again and have the following question:
Page 12 Figure 10 (and accompanying text)
Is my assumption correct here that this peak COP of the third dot has a direct relation with the set ‘on- time’ of the frequency timer (and also inductance of the coil). In other words, what you see in figure 10, is that the first 2 dots show that the magnetic field in the coil was not yet fully built up, and in the 4^th dot, the magnetic field was fully built up and the applied ‘too high voltage’ resulting in wasting energy by maintaining the magnetic field (current) longer than necessary?

Regarding “Suggestions to help achieve a Cop >1”
Page3 DGH10i1800PA diode
This diode I could not find in anywhere (google.com/rs-online.com/jinftry.com)
Page4 Coil and their numbers / “Peak spike Voltage”
-What was the outer ? of the rod that you used, do you have a link to the specs/shop?
-When you’d go to higher peak voltage, do you intend to do this with air coils? (so no iron in the core)
-When you measured the inductance of the coils in parallel, did you do that with a potmeter in series with the coils, and then a connected the coils + potmeter to a frequency generator?

In the pictures of your shared folder I saw that you used a hand coil turner. That took quite a while to turn 5x2600turns I assume?? Considering that these coils are not the last ones I need, I’m considering buying an automatic winder, incl shipping cost a Chinese one cost +/- E800,-
E.a.: https://www.ebay.nl/itm/165753660082...Bk9SR669tLaNYQ

For making your PCBs (they look great), which software did you use?
Although I probably use your PCB files when you update/upload them, I’m curious to see via YouTube how complex it is to use the software, might be a skill I’ll be picking up too..

Best regards,
Rodolphe