Reactive Power Generator (RPG) Replication Project

[Aaron]

Just downloaded Target 3001! from https://ibfriedrich.com/en/index.html. Thanks for letting me know about this. I'll start learning how to use it ASAP.

Me, Paul and RS all use Kicad for schematic and PCB design. RS uses other programs too I think - it's the only one me and Paul uses for sure. In case there is a need to be compatible with what we use in case we share project files at some point for this or other projects, I'd recommend Kicad. I keep an old 5.x version, 6 and 7 on my computer since different circuits was done with different versions. Anytime I do anything new, I only use the latest 7.0.

https://www.kicad.org/

[Donald Haas]

1. Resistance of your 16 motor coils, as in the resistance of one of them, I assume they are all about the same.

1200T OF #22 WIRE, 9.1Ω per coil. ( AVG)

2. Resistance of one of your gen coils.

600T of #20 wire, 9.5Ω (avg) The mags are 1 x 1 x 4", rectangular, air core.

3. What way are the magnets on the gen rotor facing?

N,N,N

4. If the gen coils were energized by DC, what would be the orientation of the magnetic field they produce?

Refer to pic of 6 mag gen before full assembly.

6 mag gen.jpg ???

All the leads of the coils proceed from one end and the beginnings are marked.

bro d

[Joster]

Generator looks good. What are the rise and fall times at the outputs from your gate drivers? I'm thinking it might be possible that it will be more than the ~20ns listed in the data sheet. Driving 8 of these FETs with one gate driver my be too heavy of a load. Not totally sure but worth a look. Also, wondering during normal run do the FETs or coils get hot?

[Energy_Hack]

This is a fun project and more easy to replicate than some other ideas out there. I gave it an honest try last year and spoke with many others for input. Hope somebody can get it to work as Mike and Paul have. My batteries always drained down over time. I tried what seemed like hundreds of configurations and voltages and various rotor and switch designs.
The IRFP260 MOSFET seemed to work best and stand up to the most abuse. In fact, the mosfet amp rating ended up being more important than the voltage rating as far as reliability of the switch.
Having a rotor with pick up coils may be an easier configuration than mechanically coupling the motor to a generator.
In either case power generation will drag down the rotor and cause it to draw more power.
The magic is if the batteries will stay charged up even though you’re pulling more power from them than you are returning. At best I could only measure about 1/2 the power returning, and often less than that. The batts would seem to regain their starting voltage until they got over 50% drained and then it would start falling off.
Good luck to all and please post any results.

[Donald Haas]

What are the rise and fall times at the outputs from your gate drivers?
3 pics. 1785RPM = 237.5 Hz (or so)

1785 * 8 / 60 = 238Hz

Probe is at the driver output.

1 gate pulse.jpg 2 rise time.jpg 3 fall time.jpg






Also, wondering during normal run do the FETs or coils get hot?

The fets stay cool. The coils are slightly above ambient, not warm to the touch.

bro d

[Joster]

The second pic shows the loading effect perhaps?

[Donald Haas]

Mr Hack,
Pleasure to see you sir.
I found that the IRFP260 MOSFET is as you say but I saw only about 220v spikes with them.
The sic fets that I'm using now are rated at 650v, 35A and I'm seeing over 700v spikes.
Batts are not staying up.

bro d

[Donald Haas]

Hi Joster,

Now we're getin' serious.

2nd pic shows "Miller plateau" which is beyond my knowledge.

https://electronics.stackexchange.co...pendent-on-vds

bro d

[Aaron]

The magic is if the batteries will stay charged up even though you’re pulling more power from them than you are returning. At best I could only measure about 1/2 the power returning, and often less than that. The batts would seem to regain their starting voltage until they got over 50% drained and then it would start falling off.
Good luck to all and please post any results.

How were you measuring 1/2 the power returning?

[Energy_Hack]

I used a clamp meter attached to scope and measured average current in and out of the circuit. (Current In measured as input to coils and out as line after diode).
I found the best recovery rates are with the batteries in parallel (generator - mode in Bedini books). However in this mode any recovered power goes back through the drive battery and further drains it. The original Bedini “SG”isolates the recovered power and the drive batt stays charged longer but the charging is less.
I multiplied average current by batt voltage to get average power In and Out. The power out was always less than the power in and so the batteries always drained down. I used up to (5) 100Ah batts in each bank and so they held for quite a while but would eventually die off. If the drive was pulling 5A and returning 2.5A, The bank had 40hours of capacity and so I’d leave it running for almost 20 hours before I started seeing them dropping in voltage.

[Joster]

Hi Joster,

Now we're getin' serious.

2nd pic shows "Miller plateau" which is beyond my knowledge.

https://electronics.stackexchange.co...pendent-on-vds

bro d

I'm not certain that it is, would be worth a try lifting the gate resistors for 7 of the fets and testing again. Also, in regards to charging I would simplify the charging circuit to make troubleshooting easier. Just go from the output diodes right to one of the battery banks and test charging over time.

[Donald Haas]

Hi Joster,

The batts charge better with no swapping system in place.
I've been wanting to understand how best to route the charge from diode to battery and also the neg leg.
I have Blue Sea hi amp manual switches and automotive 40amp fuses into the neg of the batts that the relays connect to on the pos side.
The rig will blow 30A fuses at startup.
I'm sure that Paul B has worked all this out for best results.
He is not making the details of his swapping and switching circuitry available.
God forbid that we are to be offered potted products.
My fuses are too slow and should probably be placed on the pos run wire from the relays.
That would take them out of the charge circuit.
It's good for me to talk about this stuff rather than work alone.

bro d

[Energy_Hack]

Bro D,

You have to cut power to the motor, swap batteries, and then turn power back on. If you try to operate the relays under load they will blowout. At least at the higher supply voltages above 24V. The guy in Holland published a schematic for doing this. (Nick Kraakman)

Pulse Motor Generator Design Considerations | Waveguide

EH

[Joster]

Donald, I sent you a private message, we should get talking, there is no reason your build shouldn't be a winner, its big enough and well made we just have to help you with the electronic side.

[Donald Haas]

Hi EH, this is my battery swapping sequence that I've programed an Arduino Mega to run and it works well with 60v banks, consistently.

2 bank swap sequence:
Turn off relays that power fet drivers and opto switching.
Turn off all relays between batts and fet boards.
5 second delay to manually turn on batt connect switches (were initially off) and give rotor a spin.
_________the above is only when Arduino is powered on.
{1. through 10. Repeat.}
1.Turn on relays for position 1
2.Turn on relays for fet drivers and opto switching.
(Motor starts or resumes running at this point)
3.Desired runtime
4.Turn off relays that power fet drivers and opto switching.
5.Turn off all relays between batts and fet boards.
6.Turn on relays for position 2
7.Turn on relays for fet drivers and opto switching.
(Motor resumes running at this point)
8.Desired runtime
9.Turn off relays that power fet drivers and opto switching.
10.Turn off all relays between batts and fet boards.
(back to 1.)

4-7 takes about 1 second.

bro d